Section 230: An Overview

April 7, 2021

Congressional Research Service https://crsreports.congress.gov R46751

SUMMARY

R46751 Section 230: An Overview April 7, 2021 Section 230 of the Communications Act of 1934, enacted as part of the Communications Decency Act of 1996, provides limited federal immunity to providers and users of interactive Valerie C. Brannon computer services. The law generally precludes providers and users from being held liable—that Legislative Attorney is, legally responsible—for information provided by a third party, but does not prevent them from being held legally responsible for information that they have developed or for activities unrelated to third-party content. Courts have interpreted Section 230 to foreclose a wide variety of lawsuits Eric N. Holmes and to preempt laws that would make providers and users liable for third-party content. For Legislative Attorney example, the law has been applied to protect online service providers like social media companies from lawsuits based on their decisions to transmit or take down user-generated content.

Two provisions of Section 230 are the primary framework for this immunity. First, Section 230(c)(1) specifies that service providers and users may not “be treated as the publisher or speaker of any information provided by another information content provider.” In Zeran v. America Online, Inc., an influential case interpreting this provision, a federal appeals court said that Section 230(c)(1) bars “lawsuits seeking to hold a service provider liable for its exercise of a publisher’s traditional editorial functions—such as deciding whether to publish, withdraw, postpone or alter content.” Second, Section 230(c)(2) states that service providers and users may not be held liable for voluntarily acting in good faith to restrict access to “obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable” material. Section 230(c)(2) is thus more limited: it applies only to good-faith takedowns of objectionable material, while courts have interpreted Section 230(c)(1) to apply to both distribution and takedown decisions.

Section 230 contains statutory exceptions. This federal immunity generally will not apply to suits brought under federal criminal law, intellectual property law, any state law “consistent” with Section 230, certain privacy laws applicable to electronic communications, or certain federal and state laws relating to sex trafficking.

In recent years, legislators and outside commentators have debated the proper scope of Section 230. While the law does have a number of defenders, others have argued that courts have interpreted Section 230 immunity too broadly. In the 116th Congress, 26 bills would have amended the scope of Section 230 immunity. These proposals ranged from outright repeal, to placing certain conditions on immunity, to creating narrower exceptions allowing certain types of lawsuits. Some bills sought to amend the scope of Section 230(c)(1), limiting “publisher” immunity in an attempt to encourage sites to take down certain types of undesirable content. Others sought to encourage sites to host more content by narrowing immunity for certain types of takedown decisions.

The executive branch also weighed in on Section 230 reform in 2020, with proposals from the National Telecommunications and Information Administration and the Department of Justice. One issue raised by these proposals concerned whether the Federal Communications Commission (FCC) has regulatory authority to implement Section 230. While the FCC generally has authority to administer the Communications Act of 1934, to date, the FCC has not played a role in interpreting or applying Section 230, and Section 230 does not explicitly mention the FCC. Commentators have thus disputed whether Congress intended to vest the FCC with regulatory authority over Section 230 and whether the statute contains any ambiguous language that could be clarified through FCC regulation.

In addition, proposals to amend Section 230 may raise two distinct types of First Amendment issues. The first issue is whether any given proposal infringes the constitutionally protected speech of either providers or users. This concern may be especially acute if a proposal restricts providers’ editorial discretion or creates content- or viewpoint-based distinctions. The second issue is whether, if Section 230 is repealed in whole or in part, the First Amendment may nonetheless prevent private parties or the government from holding providers liable for publishing content. The First Amendment might prevent some claims premised on decisions to host or restrict others’ speech, but its protections are likely less extensive than the current scope of Section 230 immunity.

Congressional Research Service Section 230: An Overview

Contents

Text and Legislative History ...... 1 Section 104: Online Family Empowerment ...... 5 Stratton Oakmont, Inc. v. Prodigy Services Co...... 6 Judicial Interpretation ...... 8 Section 230(c)(1): Publisher Activity ...... 9 Early Interpretations: Zeran v. America Online, Inc...... 10 Service Provider Role as Publisher ...... 11 Information Provided by Another Information Content Provider ...... 14 Section 230(c)(2)(A): Restricting Access to Objectionable Material...... 19 Good Faith ...... 20 Objectionable Material...... 21 Section 230(c)(2)(B): Enabling Access Restriction ...... 22 Section 230(e): Exceptions ...... 24 Federal Criminal Law ...... 24 Intellectual Property Law ...... 25 State Law ...... 27 Electronic Communications Privacy Act of 1986 ...... 28 Sex Trafficking Law (FOSTA)...... 28 Reform Proposals and Considerations for Congress ...... 29 Overview of Reform Proposals ...... 30 FCC ...... 36 Existing Legal Authority ...... 37 Deference to FCC Regulation ...... 40 Considerations for Congress ...... 42 Free Speech Considerations ...... 42 Background Principles ...... 43 First Amendment Issues with Reform Proposals ...... 47 Comparing the Operation of First Amendment and Section 230 Protections ...... 52

Contacts Author Information ...... 54

Congressional Research Service Section 230: An Overview

n 1996, Congress passed a suite of measures to amend the Communications Act of 1934 in order to protect children on the internet. The new measures were known collectively as the Communications Decency Act (CDA).1 Some portions of the CDA directly imposed liability I 2 for transmitting obscene or harassing material online, including two provisions that the Supreme Court struck down as unconstitutional in 1997.3 The CDA’s new Section 230 of the Communications Act4 took a different approach.5 It sought to allow users and providers of “interactive computer services” to make their own content moderation decisions, while still permitting liability in certain limited contexts.6 Since its passage, federal courts have interpreted Section 230 as creating expansive immunity for claims based on third-party content that appears online.7 Consequently, internet companies and users frequently rely on Section 230’s protections to avoid liability in federal and state litigation. But in recent years, commentators and jurists have expressed concern that the broad immunity courts have recognized under Section 230 is beyond the law’s intended scope.8 This report explores the origins, current application, and future of Section 230. It first discusses the history and passage of Section 230 and the CDA. The report then analyzes how courts have applied Section 230 in litigation. The report concludes with a discussion of proposed reforms to Section 230 and legal and constitutional considerations relevant to reform efforts. This report does not discuss the possible international trade implications of amending Section 230. This issue is discussed briefly in CRS Legal Sidebar LSB10484, UPDATE: Section 230 and the Executive Order on Preventing Online Censorship, by Valerie C. Brannon et al.

Text and Legislative History Congress enacted the CDA as part of the Telecommunications Act of 1996.9 According to the conference report, the CDA as a whole was intended to “modernize the existing protections

1 Pub. L. No. 104-104, Tit. V, 110 Stat. 133 (1996). 2 E.g., 47 U.S.C. § 223(d). 3 Reno v. ACLU, 521 U.S. 844, 882 (1997). 4 47 U.S.C. § 230. Although Section 230 is sometimes referred to as “Section 230 of the CDA” or “CDA Section 230,” “Section 230” more accurately refers to the statute’s place in the Communications Act. 5 141 CONG. REC. H8470 (daily ed. Aug. 4, 1995) (statement of Rep. Ron Wyden) (noting that the approach of Section 230 stands “in sharp contrast to the work of the other body,” which sought “to try to put in place the Government rather than the private sector about this task of trying to define indecent communications and protecting our kids”). 6 See 47 U.S.C. § 230(b) (expressing a deregulatory policy goal); id. § 230(e) (providing limited exceptions). 7 See, e.g., Zeran v. Am. Online, Inc. 129 F.3d 327, 330–31 (4th Cir. 1997) 8 See, e.g., Force v. Facebook, Inc., 934 F.3d 53, 84 (2d Cir. 2019) (Katzmann, J., concurring in part) (opining that Section 230 as applied creates “extensive immunity . . . for activities that were undreamt of in 1996” and “[i]t therefore may be time for Congress to reconsider the scope of § 230”); Malwarebytes, Inc. v. Enigma Software Grp. USA, LLC, 141 S. Ct. 13, 14–15 (2020) (Thomas, J., statement respecting the denial of certiorari) (positing that the “modest understanding” of what Section 230 is meant to do based on its text “is a far cry from what has prevailed in court”); 1 R. SMOLLA, LAW OF DEFAMATION § 4.86 (2d ed. 2019) (“[C]ourts have extended the immunity in § 230 far beyond anything that plausibly could have been intended by Congress.”). 9 Pub. L. No. 104-104, § 501, 110 Stat. 133–43 (1996).

Congressional Research Service 1 Section 230: An Overview

against obscene, lewd, indecent or harassing uses of a telephone.”10 Since its enactment in 1996,11 Section 230 has been amended twice: once to add a new obligation for interactive computer services to notify customers about parental control protections,12 and once to except its application in certain civil and criminal cases involving prostitution or sex trafficking.13 Section 230 contains findings14 and policy statements,15 expressing, among other things, that Congress sought to promote the free development of the internet, while also “remov[ing] disincentives” to implement “blocking and filtering technologies” that restrict “children’s access to . . . inappropriate online material”16 and “ensur[ing] vigorous enforcement of Federal criminal laws to deter and punish trafficking in obscenity, stalking, and harassment by means of computer.”17 The heart of Section 230, however, is arguably the immunity created in subsection (c):

(c) PROTECTION FOR “GOOD SAMARITAN” BLOCKING AND SCREENING OF OFFENSIVE MATERIAL.—

(1) TREATMENT OF PUBLISHER OR SPEAKER.—No provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider.

(2) CIVIL LIABILITY.—No provider or user of an interactive computer service shall be held liable on account of— (A) any action voluntarily taken in good faith to restrict access to or availability of material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally protected; or (B) any action taken to enable or make available to information content providers or others the technical means to restrict access to material described in [subparagraph (A)].18

10 S. REP. NO. 104-23, at 59 (1995); see also id. (“The decency provisions increase the penalties for obscene, indecent, harassing or other wrongful uses of telecommunications facilities; protect privacy; protect families from uninvited and unwanted cable programming which is unsuitable for children and give cable operators authority to refuse to transmit programs or portions of programs on public or leased access channels which contain obscenity, indecency, or nudity.”). The Supreme Court struck down some of these provisions as unconstitutional in Reno v. ACLU, 521 U.S. 844, 882 (1997). 11 Pub. L. No. 104-104, § 509, 110 Stat. 137–39 (1996). 12 Pub. L. No. 105-277, § 1404, 112 Stat. 2681-739 (1998). This 1998 law also amended 47 U.S.C. § 230(e)(1) to clarify that Section 230 should not be construed to impair the enforcement of 47 U.S.C. § 231, a new provision created by the 1998 law. Id. 13 Allow States and Victims to Fight Online Sex Trafficking Act of 2017 (FOSTA), Pub. L. No. 115-164, § 4, 132 Stat. 1253 (2018). FOSTA also created criminal and civil liability for owning, managing, or operating an interactive computer service “with the intent to promote or facilitate the prostitution of another person . . . .” Id. § 3. 14 47 U.S.C. § 230(a). 15 Id. § 230(b). 16 Id. § 230(b)(4). 17 Id. § 230(b)(5). 18 Id. § 230(c). Courts have read 47 U.S.C. § 230(c)(2)(B)’s reference to “paragraph (1)” to mean § 230(c)(2)(A). E.g. Zango, Inc. v. Kaspersky Lab, Inc., 568 F.3d 1169, 1173 n.5 (9th Cir. 2009) (“We take it that the reference to the ‘material described in paragraph (1)’ is a typographical error, and that instead the reference should be to paragraph (A), i.e., § 230(c)(2)(A). . . . Paragraph (1) pertains to the treatment of a publisher or speaker and has nothing to do with ‘material,’ whereas subparagraph (A) pertains to and describes material.”) (citation omitted).

Congressional Research Service 2 Section 230: An Overview

Thus, Section 230(c) contains two distinct provisions that together create a broad immunity from suit for a “provider or user of an interactive computer service.” Section 230(c)(1) specifies that service providers may not “be treated as the publisher or speaker of any information provided by 19 another information content provider,” while Section 230(c)(2) ensures that service providers may not be held liable for voluntarily acting to restrict access to objectionable material.20 Both “interactive computer service” and “information content provider” are statutorily defined terms.21 An “interactive computer service” is “any information service, system, or access software provider that provides or enables computer access by multiple users to a computer server.”22 Courts have considered online service providers such as Google,23 Yahoo!,24 Facebook,25 and Craigslist26 to be “interactive computer service” providers.27 Given the breadth of this definition, courts have also concluded that it extends to companies that provide broadband internet access28 or web hosting,29 as well as entities such as libraries30 or employers31 who provide computer access. An “information content provider” is “any person or entity that is responsible, in whole or in part, for the creation or development of information provided through the Internet or any other interactive computer service.”32 Thus, Section 230 distinguishes those who create content from those who provide access to that content, providing immunity from suit to the latter group.33 An entity may be both an “interactive computer service” provider and an “information content provider,” but the critical inquiry for applying Section 230’s immunity provisions is whether the service provider developed the content that is the basis for liability.34 Section 230(e) contains “exceptions” to the law’s immunity provision:35

19 47 U.S.C. § 230(c)(1). 20 Id. § 230(c)(2). 21 Id. § 230(f). 22 Id. § 230(f)(2). 23 E.g., Marshall’s Locksmith Serv. v. Google, LLC, 925 F.3d 1263, 1268 (D.C. Cir. 2019). 24 E.g., Barnes v. Yahoo!, Inc., 570 F.3d 1096, 1101 (9th Cir. 2009). 25 E.g., Klayman v. Zuckerberg, 753 F.3d 1354, 1357 (D.C. Cir. 2014). 26 Chi. Lawyers’ Comm. for Civil Rights Under Law, Inc. v. Craigslist, Inc., 519 F.3d 666, 671 (7th Cir. 2008). 27 See also Universal Commc’n Sys., Inc. v. Lycos, Inc., 478 F.3d 413, 419 (1st Cir. 2007) (“Providing access to the Internet is . . . not the only way to be an interactive computer service provider.”). 28 See e360Insight, LLC v. Comcast Corp., 546 F. Supp. 2d 605, 607 (N.D. Ill. 2008); see also, e.g., Winter v. Bassett, No. 1:02CV00382, 2003 WL 27382038, at *6 (M.D.N.C. Aug. 22, 2003) (concluding Section 230 protects Verizon and AT&T as interactive computer service providers). 29 Ricci v. Teamsters Union Local 456, 781 F.3d 25, 28 (2d Cir. 2015); see also, e.g., Gucci Am., Inc. v. Hall & Assocs., 135 F. Supp. 2d 409, 412 (S.D.N.Y. 2001) (describing Mindspring, a web hosting service, as an “interactive computer service”). 30 The statute specifically provides that the definition includes “such systems operated or services offered by libraries or educational institutions.” 47 U.S.C. § 230(f)(2). See, e.g., Kathleen R. v. City of Livermore, 104 Cal. Rptr. 2d 772, 777 (Cal. Ct. App. 2001) (“Respondent provides an ‘interactive computer service’ in this case because its library computers enable multiple users to access the Internet.”). 31 E.g., Miller v. Fed. Express Corp., 6 N.E.3d 1006, 1017 (Ct. App. Ind. 2014). 32 47 U.S.C. § 230(f)(3). 33 See id. § 230(c), (f). 34 See, e.g., Fair Hous. Council v. Roommates.com, LLC, 521 F.3d 1157, 1174 (9th Cir. 2008). 35 E.g., Universal Commc’n Sys., Inc. v. Lycos, Inc., 478 F.3d 413, 418 (1st Cir. 2007) (“[Plaintiff] has attempted to plead around that immunity . . . by asserting causes of action that purportedly fall into one of the statutory exceptions to Section 230 immunity.” (emphasis added)).

Congressional Research Service 3 Section 230: An Overview

(e) EFFECT ON OTHER LAWS.—

(1) NO EFFECT ON CRIMINAL LAW.—Nothing in this section shall be construed to impair the enforcement of section 223 or 231 of this title, chapter 71 (relating to obscenity) or 110 (relating to sexual exploitation of children) of title 18, United States Code, or any other Federal criminal statute.

(2) NO EFFECT ON INTELLECTUAL PROPERTY LAW.—Nothing in this section shall be construed to limit or expand any law pertaining to intellectual property.

(3) STATE LAW.—Nothing in this section shall be construed to prevent any State from enforcing any State law that is consistent with this section. No cause of action may be brought and no liability may be imposed under any State or local law that is inconsistent with this section.36

(4) NO EFFECT ON COMMUNICATIONS PRIVACY LAW.—Nothing in this section shall be construed to limit the application of the Electronic Communications Privacy Act of 1986 or any of the amendments made by such Act, or any similar State law.

(5) NO EFFECT ON SEX TRAFFICKING LAW.—Nothing in this section (other than subsection (c)(2)(A)) shall be construed to impair or limit: (A) any claim in a civil action brought under section 1595 of Title 18, if the conduct underlying the claim constitutes a violation of section 1591 of that title; (B) any charge in a criminal prosecution brought under State law if the conduct underlying the charge would constitute a violation of section 1591 of Title 18; or (C) any charge in a criminal prosecution brought under State law if the conduct underlying the charge would constitute a violation of section 2421A of Title 18, and promotion or facilitation of prostitution is illegal in the jurisdiction where the defendant's promotion or facilitation of prostitution was targeted.37 Courts have interpreted the language providing that Section 230 will not “limit” or “impair the enforcement of” other laws as creating “exceptions” to Section 230.38 As one court reasoned, if intellectual property laws would impose liability on a provider, then applying Section 230 to bar that lawsuit “would ‘limit’ the laws pertaining to intellectual property in contravention of § 230(e)(2).”39 Accordingly, Section 230 immunity generally will not apply to suits brought under federal criminal law,40 intellectual property law,41 any state law “consistent” with Section 230,42 certain electronic communications privacy laws,43 or certain federal and state laws relating to sex trafficking.44

36 In contrast to the exceptions created by most of subsection (e), courts have read the second sentence of Section 230(e)(3) to “preempt contrary state law.” E.g., Doe v. GTE Corp., 347 F.3d 655, 658 (7th Cir. 2003). 37 47 U.S.C. § 230(e). 38 See, e.g., Universal Commc’n Sys., Inc., 478 F.3d at 418. 39 Gucci Am., Inc. v. Hall & Assocs., 135 F. Supp. 2d 409, 413 (S.D.N.Y. 2001). 40 47 U.S.C. § 230(e)(1). 41 Id. § 230(e)(2). As discussed in more detail below, courts have disagreed about whether this exception includes only federal laws, or state laws as well. Infra “Intellectual Property Law.” 42 47 U.S.C. § 230(e)(3). 43 Id. § 230(e)(4). 44 Id. § 230(e)(5).

Congressional Research Service 4 Section 230: An Overview

Section 104: Online Family Empowerment Representatives Cox and Wyden offered the provision that would become Section 230 as Section 104 of House Bill 1555,45 an amendment to the House version of the CDA titled “Online Family Empowerment.”46 Representative Cox stated that Section 104 would serve two purposes: First, it will protect computer Good Samaritans, online service providers, anyone who provides a front end to the Internet, let us say, who takes steps to screen indecency and offensive material for their customers. It will protect them from taking on liability . . . . Second, it will establish as the policy of the United States that we do not wish to have content regulation by the Federal Government of what is on the Internet . . . .47 Many of those who spoke in favor of this amendment on the floor of the House argued that it would allow private parties, in the form of parents and internet service providers, to regulate offensive content, rather than the FCC.48 In particular, then-Representative Wyden emphasized that “parents and families are better suited to guard the portals of cyberspace and protect our children than our Government bureaucrats,” and argued against federal censorship of the Internet.49 The conference report echoed these concerns: This section provides “Good Samaritan” protections from civil liability for providers or users of an interactive computer service for actions to restrict or to enable restriction of access to objectionable online material. One of the specific purposes of this section is to overrule Stratton-Oakmont v. Prodigy and any other similar decisions which have treated such providers and users as publishers or speakers of content that is not their own because they have restricted access to objectionable material. The conferees believe that such decisions create serious obstacles to the important federal policy of empowering parents to

45 H.R. 1555, 104th Cong. (1995). 46 See 141 CONG. REC. H8468 (daily ed. Aug. 4, 1995). 47 See 141 CONG. REC. H8470 (daily ed. Aug. 4, 1995) (statement of Rep. Christopher Cox). See also Carafano v. Metrosplash.com, Inc., 339 F.3d 1119, 1122 (9th Cir. 2003) (“Congress enacted this provision as part of the Communications Decency Act of 1996 for two basic policy reasons: to promote the free exchange of information and ideas over the Internet and to encourage voluntary monitoring for offensive or obscene material.”); Zeran v. Am. Online, Inc., 129 F.3d 327, 330–31 (4th Cir. 1997) (“Section 230 was enacted, in part, to maintain the robust nature of Internet communication and, accordingly, to keep government interference in the medium to a minimum. . . . Another important purpose of § 230 was to encourage service providers to self-regulate the dissemination of offensive material over their services.”). 48 See 141 CONG. REC. H8470 (daily ed. Aug. 4, 1995) (statement of Rep. Christopher Cox) (“[W]e do not wish to have a Federal Computer Commission with an army of bureaucrats regulating the Internet because frankly the Internet has grown up to be what it is without that kind of help from the Government.”); id. at H8470 (statement of Rep. Joe Barton) (arguing this amendment provides “a reasonable way to . . . help [service providers] self-regulate . . . without penalty of law”); id. at H8471 (statement of Rep. Rick White) (arguing the responsibility for “protect[ing children] from the wrong influences on the Internet” should lie with parents instead of federal government); id. at H8471 (statement of Rep. Zoe Lofgren) (arguing that amendment should be adopted to “preserve . . . open systems on the Net”); id. at H8471 (statement of Rep. Bob Goodlatte) (“The Cox-Wyden amendment is a thoughtful approach to keep smut off the net without government censorship.”). Some have questioned whether the text of the amendment, in fact, prevented the federal government from regulating the Internet. See Robert Cannon, The Legislative History of Senator Exon’s Communications Decency Act: Regulating Barbarians on the Information Superhighway, 49 FED. COMM. L.J. 51, 68 (1996) (“The opposition [to the Senate version of the CDA] proclaimed that the Cox/Wyden Amendment forbade FCC regulation of the Internet; it did not. The opposition claimed that it preempted state regulation of the Internet; it did not.”) (citations omitted). 49 141 CONG. REC. H8470 (daily ed. Aug 4, 1995) (statement of Rep. Ron Wyden).

Congressional Research Service 5 Section 230: An Overview

determine the content of communications their children receive through interactive computer services.50 As originally introduced and passed by the House, Section 104 also contained a section stating that the CDA should not be construed “to grant any jurisdiction or authority” to the Federal Communications Commission (FCC) to regulate the Internet.51 However, this language was removed during the conference committee on the bill.52

Stratton Oakmont, Inc. v. Prodigy Services Co. As noted on the floor of the House53 and in the conference report,54 the amendment that would become Section 230 sought to overturn the result in Stratton Oakmont, Inc. v. Prodigy Services Co., a 1995 New York state trial court decision.55 The plaintiff in that case had sued Prodigy for libel—that is, defamation in written form.56 Although Prodigy, an internet service provider,57 had not itself made the allegedly libelous statements, the plaintiff alleged that Prodigy was legally responsible for publishing those statements because it hosted the message boards on which the statements were posted.58 Prodigy’s liability depended on a determination that the company was a “publisher,” because under ordinary principles of defamation law, a publisher like a newspaper “who repeats or otherwise republishes a libel is subject to liability as if he had originally published it.”59 By contrast, speech “distributors” such as libraries or newsstands may be held liable for circulating publications that contain defamatory statements only if they know or have reason to know of the defamatory statements.60 A 1991 decision from a federal trial court, Cubby v. CompuServe, Inc., applied this notice-based distributor liability to another early internet service provider, CompuServe, that the court determined was sufficiently similar to a newsstand.61

50 S. REP. NO. 104-230, at 194 (1996). 51 See H.R. REP. NO. 104-223, at 29 (1995); 141 CONG. REC. H8469 (daily ed. Aug. 4, 1995); 141 CONG. REC. H9988 (daily ed. Oct. 12, 1995). 52 See S. REP. NO. 104-230, at 86–87 (1996). For more information on conference committees, see CRS Report 98-696, Resolving Legislative Differences in Congress: Conference Committees and Amendments Between the Houses, by Elizabeth Rybicki. 53 141 CONG. REC. H8469–70 (daily ed. Aug. 4, 1995) (statement of Rep. Christopher Cox). 54 S. REP. NO. 104-230, at 194 (1996). 55 Stratton Oakmont, Inc. v. Prodigy Servs. Co., No. 31063/94, 1995 WL 323710 (N.Y. Sup. Ct. May 24, 1995). In contrast, Representative Cox noted approvingly a federal trial court decision holding that CompuServe could not be held liable for allegedly defamatory statements that were posted on an internet forum over which it exercised no editorial control. 141 CONG. REC. H8469 (daily ed. Aug. 4, 1995) (statement of Rep. Christopher Cox); Cubby, Inc. v. CompuServe, Inc., 776 F. Supp. 135, 140 (S.D.N.Y. 1991). 56 Stratton Oakmont, 1995 WL 323710, at *1. 57 Prodigy was “a consumer-oriented online service” that allowed users to “trade emails, participate in online message board discussions, read the daily news, shop for mail-order items, check the weather, stocks, sports scores, play games, and more.” Benj Edwards, Where Online Services Go When They Die, THE ATLANTIC (July 12, 2014), https://www.theatlantic.com/technology/archive/2014/07/where-online-services-go-when-they-die/374099. “It was very much like a microcosm of the modern Internet—if the entire World Wide Web was published by a single company.” Id. 58 See Stratton Oakmont, 1995 WL 323710, at *2. 59 Id. at *3. 60 Id. 61 Cubby, Inc. v. CompuServe, Inc., 776 F. Supp. 135, 140–41 (S.D.N.Y. 1991). See id. at 140 (“A computerized database is the functional equivalent of a more traditional news vendor, and the inconsistent application of a lower standard of liability to an electronic news distributor such as CompuServe than that which is applied to a public library,

Congressional Research Service 6 Section 230: An Overview

The plaintiffs in Stratton Oakmont argued that Prodigy should be considered a publisher rather than a distributor because it “held itself out as an online service that exercised editorial control over the content of messages posted on its computer bulletin boards.”62 Prodigy argued in response that it was more like a bookstore or newsstand than a newspaper, citing Cubby and claiming that it did not exercise “sufficient editorial control over its computer bulletin boards to render it a publisher” of the allegedly unlawful material.63 Prodigy pointed out that it did not— and could not—manually review “all messages prior to posting” them.64 The court concluded that Prodigy was a publisher of the alleged libel because it controlled the content of its message boards through an “automatic software screening program” and “Board Leaders” who removed messages that violated Prodigy’s guidelines.65 The court held that “[b]y actively utilizing technology and manpower to delete notes from its computer bulletin boards on the basis of offensiveness and ‘bad taste,’ for example, [Prodigy] is clearly making decisions as to content . . . , and such decisions constitute editorial control.”66 The court emphasized that it was Prodigy’s “conscious choice” to exercise editorial control, implemented through “policies, technology and staffing decisions,” that had “opened it up to a greater liability.”67 One of the sponsors of Section 104 argued on the floor of the House that the ruling against Prodigy was “backward.”68 Representative Cox argued that Congress should be encouraging internet service providers “like Prodigy, like CompuServe, like America Online, like the new Microsoft network, to do everything possible for us, the customer, to help us control, at the portals of our computer, at the front door of our house, what comes in and what our children see.”69 It was to this end, Representative Cox contended, that Section 104 sought to protect “computer Good Samaritans,” protecting them “from taking on liability such as occurred in the Prodigy case in New York that they should not face for helping us and for helping us solve this problem.”70 Ultimately, Section 104 made it into the CDA, largely unchanged, as Section 230.71

book store, or newsstand would impose an undue burden on the free flow of information.”). 62 Stratton Oakmont, 1995 WL 323710, at *2. 63 Id. at *3. 64 Id. 65 Id. at *4. 66 Id. (citation omitted). 67 Id. at *5. Cf. Cubby, Inc. v. CompuServe, Inc., 776 F. Supp. 135, 140 (S.D.N.Y. 1991) (“[A third party] uploads the text of Rumorville into CompuServe’s data banks and makes it available to approved . . . subscribers [to CompuServe’s publishing service] instantaneously. CompuServe has no more editorial control over such a publication than does a public library, book store, or newsstand, and it would be no more feasible for CompuServe to examine every publication it carries for potentially defamatory statements than it would be for any other distributor to do so.”); id. at 140–41 (holding CompuServe could not be held liable unless “it knew or had reason to know of the allegedly defamatory Rumorville statements”). 68 141 CONG. REC. H8470 (daily ed. Aug. 4, 1995) (statement of Rep. Christopher Cox). 69 Id. See also id. at H8471 (statement of Rep. Ron Wyden) (“Under our approach and the speed at which these technologies are advancing, the marketplace is going to give parents the tools they need . . . .”). 70 Id. at H8470 (statement of Rep. Christopher Cox). 71 See S. REP. NO. 104-230, at 86–87 (1996).

Congressional Research Service 7 Section 230: An Overview

Judicial Interpretation Courts have interpreted Section 230 as creating broad immunity that allows the early dismissal of many legal claims against interactive computer service providers,72 preempting lawsuits and statutes that would impose liability based on third-party content. Courts have generally interpreted Section 230(c)’s two separate provisions as creating two distinct liability shields. Section 230(c)(1) states that interactive computer service providers and users may not “be treated as the publisher or speaker of any information provided by another” person.73 Section 230(c)(2) provides that interactive computer service providers and users may not be “held liable” for any voluntary, “good faith” action “to restrict access to or availability of material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable.”74 One conception of these two provisions is that Section 230(c)(1) applies to claims for content that is “left up,” while Section 230(c)(2) applies to claims for content that is “taken down.”75 In practice, however, courts have also applied Section 230(c)(1) to “take down” claims, and courts sometimes collapse Section 230’s two provisions into a single liability shield or do not distinguish between the two provisions.76 A defendant’s chosen statutory basis for immunity under Section 230 is consequential: Section 230(c)(2) includes a good faith requirement absent from Section 230(c)(1), while Section 230(c)(1) is limited to claims based on another’s content.77 Section 230’s provisions apply to users and providers of “interactive computer services.”78 The statute defines an interactive computer service as “any information service, system, or access software provider that provides or enables computer access by multiple users to a computer server,” specifically mentioning services that provide access to the internet and services operated by libraries or educational institutions.79 Reviewing courts have understood this term to cover various services, including “broadband providers, hosting companies, and website operators.”80

72 See, e.g., David S. Ardia, Free Speech Savior or Shield for Scoundrels: An Empirical Study of Intermediary Immunity under Section 230 of the Communications Decency Act, 43 LOY. L.A. L. REV. 373, 438–39 (2010) (reporting that almost all unreversed federal decisions involving invocations of Section 230 between Section 230’s passage and September 30, 2009 happened at the motion to dismiss or summary judgment stage). 73 47 U.S.C. § 230(c)(1). 74 Id. § 230(c)(2). 75 E.g., Doe v. GTE Corp.. 347 F.3d 655, 659 (7th Cir. 2003); cf. Malwarebytes, Inc. v. Enigma Software Grp. USA, LLC, 141 S. Ct. 13, 15 (Thomas, J., statement respecting the denial of certiorari) (articulating this view of Section 230 before positing that “[t]his modest understanding is a far cry from what has prevailed in court”). 76 E.g., Domen v. Vimeo, Inc., No. 1:19-cv-08418, 2020 WL 217048, slip op. at *13 (S.D.N.Y. Jan. 15, 2020) (holding that 230(c)(1) and (2) both provided immunity for claims arising from video hosting provider’s decision to remove content); see Malwarebytes, 141 S. Ct. at 17 (Thomas, J., statement respecting the denial of certiorari) (collecting cases); see also Riggs v. Myspace, Inc., 444 F. App’x 986 (9th Cir. 2011) (analyzing a social media website’s decision to delete user profiles under 230(c)(1)); Ebeid v. Facebook, Inc., No. 18-CV-07030, 2019 WL 2059662, at *5 (N.D. Cal. May 9, 2019) (classifying a decision to remove posts as “publisher” conduct under 230(c)(1)). 77 Although Section 230(c)(1) refers to content created by “another information content provider,” there is not judicial agreement about whether Section 230(c)(1) applies when a plaintiff’s own content is at issue—in other words, courts are divided as to whether a plaintiff itself may be “another information content provider” under Section 230(c)(1). For more discussion of this issue, see infra note 142. 78 47 U.S.C. § 230(c). 79 Id. § 230(f)(2). 80 Jones v. Dirty World Entm’t Recordings LLC, 755 F.3d 398, 406 n.2 (6th Cir. 2014); see Ricci v. Teamsters Union Local 456, 781 F.3d 25, 27–28 (2d Cir. 2015) (noting that the definition of interactive computer service “has been construed broadly to effectuate the statute’s speech-protective purpose”); Carafano v. Metrosplash.com, Inc., 339 F.3d 1119, 1123 (9th Cir. 2003) (noting that reviewing courts have “adopt[ed] a relatively expansive definition of

Congressional Research Service 8 Section 230: An Overview

Courts have also recognized that a website operated by a print or broadcast media provider may be an interactive computer service.81 Thus, a “traditional” media outlet could receive protection under Section 230 for material posted on its website while facing a different standard for material it prints or broadcasts.82 That said, courts may deny Section 230’s protections without determining whether a party claiming its protections is a provider or user of an interactive computer service, as detailed below.83

Section 230(c)(1): Publisher Activity Section 230(c)(1) states that a provider or user of an interactive computer service will not be considered a publisher or speaker of content “provided by another information content provider.”84 Courts asked to apply Section 230(c)(1) to dismiss legal claims therefore ask three questions85: 1. Is the defendant a provider or user of an interactive computer service?86 2. Does the plaintiff seek to hold the defendant liable as a publisher or speaker? 3. Does the plaintiff’s claim arise from information provided by another information content provider? If the answer to any of these questions is “no,” Section 230(c)(1) will not bar liability. As discussed above, courts have construed the definition of “interactive computer service” broadly.87 Cases thus often turn on the answers to the other two questions, which depend on the legal claims’ specific facts: an entity may act as an information content provider for certain content, but still be entitled to protection under Section 230(c)(1) for other content.88 This section will first summarize Section 230(c)(1) case law before probing specific judicial interpretations of

‘interactive computer service’”); IAN C. BALLON, 4 E-COMMERCE & INTERNET LAW 37.05[2] (2020 update) (“[A]lmost any networked computer service would qualify as an interactive computer service, as would an access software provider.”). But see FTC v. LeadClick Media, LLC, 838 F.3d 158, 175 (2d Cir. 2016) (expressing doubt that a marketing company that tracked website traffic was an interactive computer service). 81 See, e.g., Straw v. Streamwood Chamber of Commerce, Inc., No. 1-14-3094, 2015 IL App (1st) 143094-U, at *8 (Ill. App. Ct. Sept. 29, 2015) (applying Section 230 to a letter to the editor published on a newspaper’s website). 82 Cf. Blumenthal v. Drudge, 992 F. Supp. 44, 49 (D.D.C. 1998) (“Congress decided not to treat providers of interactive computer services like other information providers such as newspapers, magazines or television and radio stations[.]”). 83 See, e.g., LeadClick Media, 838 F.3d at 176 (determining that claims against defendant were based on information developed by defendant); FTC v. Accusearch, Inc., 570 F.3d 1187, 1197–98 (reaching the same conclusion and choosing to leave the question of whether defendant is an interactive computer service “to another day”). 84 47 U.S.C. § 230(c)(1). 85 See, e.g., Universal Commc’n Sys., Inc. v. Lycos, Inc., 478 F.3d 413, 418 (1st Cir. 2007); Jones v. Dirty World Entm’t Recordings LLC, 755 F.3d 398, 409 (6th Cir. 2014). 86 Although many cases involving Section 230(c)(1) are brought against providers of interactive computer services, Section 230(c)(1) also provides protection to users of interactive computer services. See, e.g., Barrett v. Rosenthal, 146 P.3d 510, 526–27 (Cal. 2006) (applying Section 230(c)(1) to an individual who posted a third-party article on a message board); see also Batzel v. Smith, 333 F.3d 1018, 1031 (9th Cir. 2003) (opining that a website’s operator is a “user” of interactive computer services, such as internet access service, and is therefore entitled to protection under Section 230(c)(1)). 87 See supra “Text and Legislative History.” 88 See, e.g., Fair Hous. Council v. Roommates.com, LLC, 521 F.3d 1157, 1162 (9th Cir. 2008) (observing that a website may avoid liability under Section 230(c)(1) for “passively display[ing] content that is created by third parties,” but such website could be subject to liability for “content that it creates itself”).

Congressional Research Service 9 Section 230: An Overview

when a service provider is acting as a publisher of another’s information or an information content provider.

Early Interpretations: Zeran v. America Online, Inc. While the legislative history of Section 230 reflects, among other things, an intent to overturn the result in Stratton Oakmont, courts have applied Section 230(c)(1) broadly to cover other circumstances. The first federal court of appeals decision to examine the scope of Section 230(c)(1) was the Fourth Circuit’s 1997 decision in Zeran v. America Online, Inc.,89 a case with several differences from Stratton Oakmont. Since its publication, other courts of appeals have largely adopted Zeran’s reasoning and broadly construed Section 230(c)(1).90 In Zeran, an unidentified user on an America Online (AOL) bulletin board posted an advertisement for T-shirts featuring slogans celebrating the bombing of the Alfred P. Murrah Federal Building in Oklahoma City.91 The user invited AOL subscribers interested in purchasing these shirts to call the plaintiff, Kenneth Zeran, at his home phone number and “ask for Ken” upon calling.92 Despite this invitation, Zeran did not post the ad himself, nor did he direct anyone to post the ad on his behalf.93 Zeran received harassing and threatening calls, and consequently he contacted AOL and asked the company to remove the ad.94 An AOL employee assured Zeran that AOL would take down the ad, but after AOL removed the ad, a similar ad took its place.95 Zeran brought negligence claims against AOL on the theory that once Zeran notified AOL of the ads, AOL had a duty to remove the ads, notify users that the ads were deceptive, and screen for similar postings.96 Zeran premised his claim against AOL on a theory of “distributor” liability.97 At common law, as discussed above, 98 vendors and distributors of defamatory publications are liable for the content of those publications if they know or have reason to know of the illegal or tortious content.99 Central to Zeran’s theory was the notion that, although Section 230(c)(1) prohibited the court

89 Zeran v. Am. Online, Inc., 129 F.3d 327 (4th Cir. 1997). For purposes of brevity, references to a particular circuit in this memorandum (e.g., the Fourth Circuit) refer to the U.S. Court of Appeals for that particular circuit (e.g., the U.S. Court of Appeals for the Fourth Circuit). 90 See Ben Ezra, Weinstein, & Co., Inc. v. Am. Online, Inc., 206 F.3d 980 (10th Cir. 2000); Green v. Am. Online (AOL), 318 F.3d 465 (3d Cir. 2003); Batzel v. Smith, 333 F.3d 1018 (9th Cir. 2003); Universal Commc’n Sys., Inc. v. Lycos, Inc., 478 F.3d 413 (1st Cir. 2007); Doe v. MySpace, Inc., 528 F.3d 413 (5th Cir. 2008); Johnson v. Arden, 614 F.3d 785 (8th Cir. 2010); Klayman v. Zuckerberg, 753 F.3d 1354 (D.C. Cir. 2014); Jones v. Dirty World Entm’t Recordings LLC, 755 F.3d 398 (6th Cir. 2014); Ricci v. Teamsters Union Local 456, 781 F.3d 25 (2d Cir. 2015); see also Almeida v. Amazon.com, Inc., 456 F.3d 1316 (11th Cir. 2006) (recognizing agreement among other courts of appeals but reaching a decision on other grounds); cf. Chi. Lawyers’ Comm. for Civil Rights Under Law, Inc. v. Craigslist, Inc., 519 F.3d 666 (7th Cir. 2008) (partially rejecting the reasoning in Zeran but nonetheless finding that Section 230 barred Fair Housing Act claims against online service provider). 91 Zeran, 129 F.3d at 329. 92 Id. 93 Zeran v. Am. Online, Inc., 958 F. Supp. 1124, 1126 (E.D. Va. 1997). 94 Zeran, 129 F.3d at 329. 95 Id. 96 Id. at 330. 97 Though Zeran characterized his claims as stemming from America Online’s negligence, the Fourth Circuit noted that the claims were “indistinguishable from a garden variety defamation action.” Id. at 332. 98 See supra “Stratton Oakmont, Inc. v. Prodigy Services Co.” 99 See Cubby, Inc. v. CompuServe, Inc., 776 F. Supp. 135, 139–40 (S.D.N.Y. 1991). This limitation on distributor liability is rooted in the First Amendment. Id. (citing Smith v. California, 361 U.S. 147, 152–53 (1959)).

Congressional Research Service 10 Section 230: An Overview

from holding AOL liable as a “publisher” of the defamatory statements, as the court treated Prodigy in Stratton Oakmont, 100 it did not eliminate notice-based distributor liability. In support of this argument, Zeran noted that Section 230 specifically uses the term “publisher.”101 The Fourth Circuit rejected this argument. Writing for a unanimous panel, Chief Judge Wilkinson posited that “distributor” liability depends on a distributor’s publication of tortious material, and a distributor is therefore a publisher.102 Judge Wilkinson therefore reasoned that both at common law and in Section 230, the use of the term “publisher” includes original publishers as well as distributors.103 The court suggested that subjecting a computer service provider to liability based on the provider’s knowledge would “reinforce[] service providers’ incentives to restrict speech and abstain from self-regulation” and “deter service providers from regulating the dissemination of offensive material over their own services.”104 Chief Judge Wilkinson therefore concluded that reading Section 230(c)(1) to leave notice-based distributor liability intact would conflict with Section 230’s purposes.105 As discussed below, Zeran has informed the approach of a vast number of courts interpreting Section 230(c)(1). As one commentator has noted, “the rule of Zeran [barring distributor liability] has been uniformly applied by every federal circuit court to consider it and by numerous state courts.”106 Even so, some jurists have expressed skepticism with the Fourth Circuit’s approach. In a statement written to accompany a denial of certiorari in a Section 230 case, U.S. Supreme Court Justice Clarence Thomas suggested, contrary to the holding in Zeran, that Section 230(c)(1) might not limit distributor liability.107

Service Provider Role as Publisher While Zeran may be understood as addressing Section 230(c)(1)’s general scope, the case also addressed how courts may determine whether a defendant claiming protection under Section 230(c)(1) should be treated as a “publisher or speaker” of another’s content.108 The Zeran court determined that the provision bars “lawsuits seeking to hold a service provider liable for its exercise of a publisher’s traditional editorial functions—such as deciding whether to publish, withdraw, postpone, or alter content.”109 More generally, the Fourth Circuit interpreted Section 230(c)(1) as “creat[ing] a federal immunity to any cause of action that would make service providers liable for information originating with a third-party user of the service.”110 This

100 Stratton Oakmont, Inc. v. Prodigy Servs. Co., No. 31063/94, 1995 WL 323710, at *4 (N.Y. Sup. Ct. May 24, 1995). 101 See Zeran, 129 F.3d at 331–32. 102 Id. at 332 (citing W. PAGE KEETON ET AL., PROSSER AND KEETON ON THE LAW OF TORTS § 113, at 803 (5th ed. 1984)). 103 Id. at 333–34. 104 Id. at 333. 105 Id. 106 Ian C. Ballon, Zeran v. AOL and Its Inconsistent Legacy, LAW JOURNAL NEWSLETTERS (Dec. 2017), https://www.lawjournalnewsletters.com/sites/lawjournalnewsletters/2017/12/01/zeran-v-aol-and-its-inconsistent- legacy/?slreturn=20201103124726 (noting, though, that different federal appeals courts apply Zeran differently). 107 Malwarebytes, Inc. v. Enigma Software Grp. USA, LLC, 141 S. Ct. 13, 15–16 (2020) (Thomas, J., statement respecting the denial of certiorari) (arguing that the imposition of distributor liability elsewhere in the CDA and the use of terms different from those used in Stratton Oakmont might suggest that Section 230 was not meant to limit distributor liability). 108 See generally Force v. Facebook, Inc., 934 F.3d 53, 64 n.18 (2d Cir. 2019) (discussing the scope of “publisher liability”). 109 Zeran, 129 F.3d at 330. 110 Id.

Congressional Research Service 11 Section 230: An Overview

interpretation would apply beyond the defamation claims brought in Zeran and Stratton Oakmont, and courts of appeals have barred many claims on the theory that the defendant computer service is being treated as a publisher or speaker.111 Many courts have used the Zeran court’s description of “traditional editorial functions”112 to determine whether a claim would impermissibly treat a service provider or user as a publisher or speaker of another’s content.113 For instance, courts have held that Section 230(c)(1) barred several claims seeking to hold sites liable for failing to adopt safety features or content policies that plaintiffs claim would have prevented violence.114 To take one example, in Doe v. MySpace, the Fifth Circuit affirmed the dismissal of a lawsuit alleging that MySpace acted negligently in failing “to implement basic safety measures to prevent sexual predators from communicating with minors on its Web site.”115 The plaintiff, a minor, had used the site to meet and communicate with an older teenager who later sexually assaulted her at an in-person meeting.116 The plaintiff argued that her negligence claims depended on “MySpace’s failure to implement basic safety measures” and therefore would not treat the site as a publisher.117 The Fifth Circuit disagreed, saying the allegations were “merely another way of claiming that MySpace was liable for publishing the communications.”118 In the court’s view, the negligence claims hinged on MySpace’s publisher functions: its decisions relating to the “monitoring, screening, and deletion” of third-party content.119 As a result, Section 230(c)(1) barred liability.120 A number of courts have held that Section 230 not only bars lawsuits seeking monetary damages, but also bars suits for injunctive relief that would require sites to take specific actions with respect to third-party content.121 For example, in Hassell v. Bird, the California Supreme Court said that

111 See, e.g., Doe v. Backpage.com, LLC, 817 F.3d 12, 18–24 (1st Cir. 2016) (applying Section 230(c)(1) to claims brought under federal and state sex trafficking statutes); Doe v. MySpace, Inc., 528 F.3d 413, 420 (5th Cir. 2008) (rejecting negligence liability for a service provider when an adult user used the service to meet and allegedly abuse minor children); Chi. Lawyers’ Comm. for Civil Rights Under Law, Inc. v. Craigslist, Inc., 519 F.3d 666, 668–69 (7th Cir. 2008) (affirming dismissal of a federal housing discrimination claim); Force v. Facebook, Inc.., 934 F.3d 53, 65– 68 (2d Cir. 2019) (applying Section 230(c)(1) to federal civil claims based on terrorist attacks encouraged and coordinated by users of a service); Universal Commc’n Sys., Inc. v. Lycos, Inc., 478 F.3d 413, 422 (1st Cir. 2007) (affirming dismissal of claims brought under state securities and cyberstalking laws). 112 Zeran, 129 F.3d at 330. 113 See, e.g., Hassell v. Bird, 420 P.3d 776, 789 (Cal. 2018); Jones v. Dirty World Entm’t Recordings LLC, 755 F.3d 398, 407 (6th Cir. 2014); Barnes v. Yahoo! Inc., 570 F.3d 1096, 1102 (9th Cir. 2009). 114 See, e.g., Herrick v. Grindr LLC, 765 Fed. App’x 586, 590–91 (2d Cir. 2019) (affirming dismissal of product liability, negligence, and infliction of emotional distress claims alleging Grindr should have adopted safety features that would have prevented a user from an ex-boyfriend’s “campaign of harassment” conducted on the service); Klayman v. Zuckerberg, 753 F.3d 1354, 1355 (D.C. Cir. 2014) (affirming dismissal of assault and negligence claims alleging Facebook should have removed “Third Palestinian Intifada” page). 115 Doe v. MySpace, 528 F.3d 413, 416 (5th Cir. 2008). 116 Id. The suit was brought by the minor and her mother under the aliases Jane and Julie Doe. See id. at 415–16. This report refers to a singular plaintiff for convenience. 117 Id. at 419. 118 Id. at 420. 119 See id. (quoting Green v. Am. Online (AOL), 318 F.3d 465, 471 (3rd Cir. 2003)). 120 Id. at 422. 121 See, e.g., Hassell v. Bird, 420 P.3d 776, 788 (Cal. 2018) (plurality opinion); id. at 794 (Kruger, J., concurring); Giordano v. Romeo, 76 So. 3d 1100, 1102 (Fla. Dist. Ct. App. 2011); Noah v. AOL Time Warner Inc., 261 F. Supp. 2d 532, 540 (E.D. Va. 2003). See also Noah, 261 F. Supp. 2d at 539–40 (collecting Section 230 cases dismissing claims for injunctive relief and concluding that the “continuing authority” of a 1998 trial court case holding that Section 230 did not bar injunctive relief was “questionable”).

Congressional Research Service 12 Section 230: An Overview

Section 230 required the dismissal of a claim that sought to enforce a court order against Yelp.122 The plaintiffs had sued the author of allegedly defamatory statements posted about their business on Yelp, and obtained a default judgment in their favor after the defendant failed to respond to the lawsuit.123 The plaintiff then attempted to enforce that judgment against Yelp, who was not originally a party to the litigation, asking the court to enter an injunction requiring Yelp to take down the defamatory statements.124 In the state court’s view, the lawsuit sought “to overrule Yelp’s decision to publish the three challenged reviews,” impermissibly treating it as a publisher of third-party information.125 The court said that allowing injunctions could “impose substantial burdens” on internet intermediaries, contrary to Section 230’s goal of “spar[ing] republishers of online content . . . from this sort of ongoing entanglement with the courts.”126 In limited circumstances, courts have concluded that a particular claim does not treat a defendant as a publisher or speaker and is thus not barred by Section 230. One such case involved a negligent failure to warn claim against a service provider, arguing that under state law, the provider had a duty to warn users that third parties had used its site to target and lure victims in a “rape scheme.”127 The court held that Section 230 did not bar the claim because the alleged duty resulted from information the service provider acquired offline, rather than from user content generated on the provider’s website, and the service provider could satisfy this duty to warn without removing any user content or changing how it monitored user content.128 Similarly, the Ninth Circuit refused to bar a state contract law claim based on a provider’s promise to remove third-party content.129 The court said that liability for the “promissory estoppel” claim came “not from [the provider’s] publishing conduct, but from [the provider’s] manifest intention to be legally obligated to do something, which happens to be removal of material from publication.”130 Claims founded on economic regulations of online services have also survived Section 230(c)(1) preemption. For example, in City of Chicago v. Stubhub!, Inc., the Seventh Circuit declined to apply Section 230(c)(1) to bar collection of a city amusement tax from an online ticket resale platform, noting that the tax “does not depend on who ‘publishes’ any information or is a ‘speaker.’”131 Similarly, the Ninth Circuit held that Section 230(c)(1) did not preempt a local ordinance regulating short-term property rentals, as applied to websites that hosted listings of such rentals.132 Among other provisions, the ordinance prohibited hosting platforms “from

122 Hassell, 420 P.3d at 778–79 (plurality opinion); id. at 794 (Kruger, J., concurring). 123 Id. at 780–81 (plurality opinion). 124 Id. at 781–82. 125 Id. at 789; accord id. at 794 (Kruger, J., concurring). See also id. at 790 (plurality opinion) (“The duty that plaintiffs would impose on Yelp, in all material respects, wholly owes to and coincides with the company's continuing role as a publisher of third party online content.”). 126 Id. at 791 (plurality opinion). See also Noah, 261 F. Supp. 2d at 540 (“[G]iven that the purpose of § 230 is to shield service providers from legal responsibility for the statements of third parties, § 230 should not be read to permit claims that request only injunctive relief. After all, in some circumstances injunctive relief will be at least as burdensome to the service provider as damages, and is typically more intrusive.”). 127 Doe v. Internet Brands, Inc., 824 F.3d 846, 849 (9th Cir. 2016). 128 Id. at 851. 129 Barnes v. Yahoo! Inc., 570 F.3d 1096, 1107 (9th Cir. 2009). 130 Id. See also, e.g., Darnaa, LLC v. Google, Inc., No. 15-cv-03221-RMW, 2016 WL 6540452, at *8 (N.D. Cal. Nov. 2, 2016) (“Plaintiff’s claim for breach of the implied covenant of good faith and fair dealing . . . is not precluded by § 230(c)(1) because it seeks to hold defendants liable for breach of defendants’ good faith contractual obligation to plaintiff, rather than defendants’ publisher status.”). 131 City of Chicago v. Stubhub!, Inc., 624 F.3d 363, 366 (7th Cir. 2010). 132 HomeAway.com, Inc. v. City of Santa Monica, 918 F.3d 676 (9th Cir. 2019).

Congressional Research Service 13 Section 230: An Overview

completing any booking transaction for properties not licensed and listed” in a city-run registry of rental properties.133 In the Ninth Circuit’s view, the ordinance merely required platforms to monitor booking transactions and did not require platforms to police the content of third-party listings.134 The court thus did not believe that the ordinance would impermissibly treat the platforms as publishers of third-party content.135 Federal courts have also declined to apply Section 230(c)(1) to lawsuits brought by the Federal Trade Commission (FTC) against service providers alleging violations of Section 5 of the Federal Trade Commission Act.136 The first Court of Appeals to address this issue was the Tenth Circuit in FTC v. Accusearch, Inc.137 The majority opinion in Accusearch did not decide whether the defendant was being treated as a publisher or speaker, instead concluding that Section 230 did not bar the suit because the defendant had contributed to the allegedly unlawful content.138 However, Judge Tymkovich wrote in a concurring opinion that the cause of action sought to hold the defendant liable for its own conduct, rather than for third-party content, and thus the defendant was not being treated as a publisher or speaker.139 In FTC v. Leadclick Media, LLC, the Second Circuit agreed with Judge Tymkovich’s concurrence and determined that a claim brought under Section 5 of the FTC Act depended on the defendant’s own deceptive acts or practices and therefore did not treat the defendant as a publisher or speaker.140

Information Provided by Another Information Content Provider Section 230(c)(1)’s protections extend only to claims that would hold a defendant liable for “information provided by another information content provider.”141 Put another way, Section 230(c)(1) does not protect defendants from claims arising from their own content.142 For example, Section 230(c)(1) would not bar a defamation claim against a social media website based on the

133 Id. at 680. 134 Id. at 682. 135 Id. at 682–83. Cf. Airbnb, Inc. v. City of Boston, 386 F. Supp. 3d 113, 120–24 (D. Mass. 2019) (ruling that a similar regulation was not preempted by Section 230, but concluding Section 230 likely did preempt portions of the regulation requiring a “booking agent” to remove improper listings). 136 15 U.S.C. § 45. 137 FTC v. Accusearch, Inc., 570 F.3d 1187, 1197 (10th Cir. 2009). 138 Id. 139 Id. at 1204 (Tymkovich, J., concurring). For more discussion of Accusearch, see infra “Subsequent Developments in Material Contribution Analysis.” 140 FTC v. Leadclick Media, LLC, 838 F.3d 158, 176–77 (2d Cir. 2016). 141 47 U.S.C. § 230(c)(1) (emphasis added). 142 A separate but related question is whether a plaintiff bringing claims based on their own content is “another information content provider” under Section 230(c)(1). Some courts have declined to apply Section 230(c)(1) to content created by a plaintiff, reasoning that allowing Section 230(c)(1) to cover such content would render Section 230(c)(2) superfluous. See, e.g., e-ventures Worldwide, LLC v. Google, Inc., No. 2:14-cv-646-FtM-PAM-CM, 2017 WL 2210029, at *3 (M.D. Fl. Feb. 8, 2017) (declining to apply Section 230(c)(1) to unfair competition claims based on Google’s removal of plaintiff’s advertising material). Other courts have applied Section 230(c)(1) to such claims. See, e.g., Riggs v. MySpace, Inc., 444 F. App’x 986, 987 (9th Cir. 2011) (affirming dismissal under Section 230(c)(1) of claims based on removal of plaintiff-created profile pages); Sikhs for Justice “SFJ”, Inc. v. Facebook, Inc., 144 F. Supp. 3d 1088, 1093–94 (N.D. Cal. 2015) (applying Section 230(c)(1) to dismiss claims based on blocking access to plaintiff-created page); cf. Batzel v. Smith, 333 F.3d 1018, 1031 (9th Cir. 2003) (interpreting Section 230(c)(1)’s reference to “another information content provider” to “distinguish[] the circumstance in which the interactive computer service itself meets the definition of ‘information content provider’ with respect to the information in question”).

Congressional Research Service 14 Section 230: An Overview

content of a label or disclaimer added by the website to third-party content.143 But as recognized in Zeran and other cases, Section 230(c)(1) does allow a defendant to make some editorial adjustments to third-party content without being considered the provider of that content.144 Whether a defendant is being treated as the publisher of information provided by “another information content provider” depends in part on whether the defendant is an information content provider itself.145 As defined in Section 230, an “information content provider” is “any person or entity that is responsible, in whole or in part, for the creation or development of information provided through the Internet or any other interactive computer service.”146 When a case involves third-party content, courts routinely focus on the defendant’s role in the “creation or development” of the content.147

Fair Housing Council v. Roommates.com, LLC A foundational case on this issue is the Ninth Circuit’s decision in Fair Housing Council v. Roommates.com, LLC (Roommates).148 In Roommates, housing agencies in San Diego and the San Fernando Valley sued the operators of the website Roommates.com,149 a website that allows individuals to locate prospective roommates.150 New Roommates.com users were required to complete a questionnaire that included the user’s preferences for a roommate’s age, gender, sexual orientation, and number of children.151 Roommates.com then displayed the answers to these questions in personal profiles, which users of the site could search and view.152 The housing agencies alleged that Roommates.com had violated a provision of the Fair Housing Act that prohibits publishing advertisements for the sale or rental of a dwelling that indicate any preference based on sex, familial status, or other protected characteristics.153 In defense, Roommates.com argued that the housing agencies were seeking to hold Roommates.com liable for content generated by individual users and therefore Section 230(c)(1) would bar liability.154 In

143 Cf. Maffick, LLC v. Facebook, Inc., No. 20-05222, 2020 WL 5257853 (N.D. Cal. Sept. 3, 2020) (ignoring Section 230 entirely in a case based on Facebook’s labeling of user accounts as “Russia state-controlled media”); see also, e.g., Maynard v. Snapchat, 816 S.E.2d 77, 79 (Ga. Ct. App. 2018) (concluding Section 230 did not bar negligence claim premised on Snapchat’s “Speed Filter,” which superimposed the user’s speed over user pictures or videos). 144 Zeran v. Am. Online, Inc., 129 F.3d 327, 330 (4th Cir. 1997); see Batzel v. Smith, 333 F.3d 1018, 1031 (9th Cir. 2003) (making minor alterations to email before posting email to listserv did not render defendant liable for third-party content); Ben Ezra, Weinstein, & Co. v. America Online, Inc., 206 F.3d 980, 985–86 (10th Cir. 2000) (deleting erroneous information from a database containing third-party content did not render defendant liable for third-party content); Blumenthal v. Drudge, 992 F. Supp. 44, 51–52 (D.D.C. 1998) (reserving right to “require reasonable changes” to content did not render service provider liable for content). 145 See 47 U.S.C. § 230(c)(1). 146 Id. § 230(f)(3). 147 See, e.g., Batzel, 333 F.3d at 1031; Ben Ezra, Weinstein, & Co., 206 F.3d at 985. 148 Fair Hous. Council v. Roommates.com, LLC, 521 F.3d 1157 (9th Cir. 2008). 149 The defendant’s corporate name in Roommates is the singular Roommate.com, LLC. However, the domain of the website operated by the defendant is the plural roommates.com. This linguistic mismatch resulted in the party being named as “Roommates.com” in the Ninth Circuit case. Cf. Fair Hous. Council v. Roommate.com, LLC, No. 03-09386, 2004 WL 3799488 (C.D. Cal. Sept. 30, 2004). For clarity, this report will refer to the defendant website operator as “Roommates.com.” 150 Roommates, 521 F.3d at 1162. 151 Id. at 1161. 152 Id. 153 42 U.S.C. § 3604(c). 154 Roommates, 521 F.3d at 1162.

Congressional Research Service 15 Section 230: An Overview

an en banc rehearing, the Ninth Circuit rejected this contention, saying that Roommates.com’s required questionnaire “induce[d] third parties to express illegal preferences.”155 According to the court, because this questionnaire was created by Roommates.com and not its users, Section 230(c)(1) did not apply.156 Addressing Roommates.com’s liability for displaying its user’s preferences on personal profiles, the court acknowledged that the “illegal preferences” at issue were pieces of information provided by information content providers other than Roommates.com.157 But the Ninth Circuit noted that Roommates.com may still have “develop[ed] . . . in part” this information, such that Roommates.com could be considered the “information content provider” of the information.158 The court determined that by requiring users to answer its questionnaire, Roommates.com had at least in part developed the information.159 The Ninth Circuit cabined the reach of its holding by specifying that “passive conduits” or “neutral tools,” such as a search engine that filters content only by user-generated criteria, would not be responsible for developing content, and that “development” as used in Section 230 means “materially contributing to its alleged unlawfulness.”160 The court also concluded that Section 230(c)(1) did bar liability for user comments made in an “Additional Comments” section of user profiles, a blank box where users could post text with no constraints.161 Writing for the majority, Chief Judge Kozinski summarized the Roommates court’s holding: “a website helps to develop unlawful content . . . if it contributes materially to the alleged illegality of the conduct.”162 In a later Ninth Circuit opinion, the court clarified that this “material contribution” test “draw[s] the line at ‘the crucial distinction between, on the one hand, taking actions (traditional to publishers) that are necessary to the display of unwelcome and actionable content and, on the other hand, responsibility for what makes the displayed content illegal or actionable.’”163

Subsequent Developments in Material Contribution Analysis Since the Ninth Circuit’s decision in Roommates, other federal courts of appeals have adopted variations on the Roommates “material contribution” analysis in determining whether a defendant is the information content provider of the information at issue. The next federal appeals court to consider Roommates was the Tenth Circuit in FTC v. Accusearch, Inc., which adopted—and possibly expanded upon—the Ninth Circuit’s reasoning in Roommates.164 At issue in Accusearch was whether a website that sold information contained in telephone records could claim Section 230 protection from an FTC enforcement action when the operator acquired these records from

155 Id. at 1165. 156 Id. 157 Id. 158 Id.; see 47 U.S.C. § 230(c)(1) (applying only to information provided by “another information content provider”). 159 Roommates, 521 F.3d at 1166. 160 Id. at 1167–69. 161 Id. at 1173–75; see also Chi. Lawyers’ Comm. for Civil Rights Under Law v. Craigslist, Inc., 519 F.3d 666, 671 (2008) (concluding Section 230(c)(1) barred a similar Fair Housing Act case brought against website that hosted apartment listings, but listings were written entirely by users). 162 Roommates, 521 F.3d at 1168. 163 Kimzey v. Yelp! Inc., 836 F.3d 1263, 1269 n.4 (9th Cir. 2016) (quoting Jones v. Dirty World Entmt. Recordings LLC, 755 F.3d 398, 413–14 (6th Cir. 2014). 164 FTC v. Accusearch, Inc., 570 F.3d 1187 (10th Cir. 2009).

Congressional Research Service 16 Section 230: An Overview

third parties.165 Accusearch argued that it did not add anything to the information after receiving it and thus was not an information content provider of the information.166 In an opinion written by Judge Hartz, the Tenth Circuit held that a defendant’s solicitation of and payment for telephone records rendered the defendant an information content provider of these records.167 The Tenth Circuit focused on whether the defendant had played any role in “developing” the information. Judge Hartz opined that the inclusion of two terms—“creation” and “development”—in Section 230’s definition of “information content provider” suggested that the two terms had distinct meanings.168 Unwilling to adopt a redundant definition of “development,” the court turned to dictionary definitions of the term and determined that information may be “developed” when the information is made “‘visible,’ ‘active,’ or ‘usable.’”169 The Tenth Circuit therefore concluded that by making telephone records public on its website, the defendant had “developed” those records.170 Noting that Section 230 defines an information content provider as one “responsible, in whole or in part” for the creation or development of content,171 the Accusearch court followed Roommates in holding that a party is “responsible” for content only when the party “in some way specifically encourages development of what is offensive about the content.”172 To the Tenth Circuit, what was “offensive” about the information at issue was that it had been publicly exposed: as the court observed, federal law generally prohibits the disclosure of telephone records to third parties.173 Judge Hartz noted that Accusearch had “affirmatively solicited” telephone records from its paid researchers and “knowingly sought to transform virtually unknown information into a publicly available commodity,” and was therefore responsible for the records being made public.174 Courts interpreting Roommates and Accusearch have attempted to define the contours of when a defendant has or has not “materially contributed” to content. A North Carolina appellate court held that “a website must effectively control the content . . . or take other actions which essentially ensure the creation of unlawful content” to be considered an information content provider.175 The Sixth Circuit has emphasized that mere encouragement does not rise to the level of material contribution, asserting that holding otherwise “would inflate the meaning of ‘development’ to the point of eclipsing the immunity from publisher-liability that Congress established.”176 Even the Ninth Circuit has cautioned against broad application of Roommates, declining to hold, for example, that a defendant materially contributed to content when the

165 Id. at 1190. 166 Id. at 1197–98. 167 Id. at 1200. 168 Id. at 1198. 169 Id. (quoting WEBSTER’S THIRD NEW INT’L DICTIONARY 618 (2002)). 170 Id. 171 47 U.S.C. § 230(f)(3). 172 Accusearch, 570 F.3d at 1199. 173 Id.; see 47 U.S.C. § 222. 174 Accusearch, 570 F.3d at 1200. 175 Hill v. Stubhub, Inc., 727 S.E.2d 550, 561 (N.C. App. 2012). 176 Jones v. Dirty World Entmt. Recordings LLC, 755 F.3d 398, 414 (6th Cir. 2014); see Fair Hous. Council v. Roommates.com, LLC, 521 F.3d 1157, 1161 n.19 (9th Cir. 2008) (noting that Roommates.com “does much more than encourage or solicit”).

Congressional Research Service 17 Section 230: An Overview

defendant did not “require[] users to post specific content,” as Roommates.com did by requiring users to complete its questionnaire.177

Algorithmic Sorting A recurring issue in Section 230 cases is whether Section 230(c)(1) immunizes the use of algorithms to filter and sort content in a particular way—a common feature on social media websites and search engines. Claims brought against websites for their use of algorithms often cast a website’s use of algorithms either as “development” of third-party content, much like the theories of Roommates and Accusearch, or as nonpublisher activity to which Section 230(c)(1) would not apply. Federal courts of appeals that have considered this issue thus far have uniformly rejected these theories.178 A thorough examination of the relationship between algorithmic content and Section 230 is the Second Circuit’s opinion in Force v. Facebook, a case brought by victims of terrorist attacks allegedly coordinated and encouraged on Facebook by individual users.179 In Force, the plaintiffs contended that Facebook’s use of algorithms to display personalized content and friend suggestions was nonpublisher activity outside Section 230’s scope or, alternatively, materially contributed to the development of user content by “mak[ing] that content more visible, available, and usable.”180 The Second Circuit declined to endorse either of these arguments and instead held that Section 230 barred the plaintiffs’ claims.181 Addressing the first argument, the court noted that how and where to display content is a quintessential editorial decision protected under Section 230, and therefore plaintiffs sought to hold Facebook liable as a publisher.182 The Second Circuit likewise held that Facebook had not developed user content when its algorithms “take the information provided by Facebook users and ‘match’ it to other users—again, materially unaltered—based on objective factors applicable to any content.”183 The Force court’s treatment of algorithmic sorting applies the “neutral tools” language first appearing in Roommates.184 Several earlier cases adopt a similar approach to such neutral tools that, though originating with this language from Roommates, slightly diverges from Roommates’ material contribution analysis. In an early case on the issue, the D.C. Circuit held that “a website

177 Dyroff v. Ultimate Software Grp., Inc., 934 F.3d 1093, 1099 (9th Cir. 2019). 178 E.g., id. at 1098–99 (opining that plaintiffs could not frame “website features as content” and that the site’s recommendation and notification functions did not materially contribute to alleged unlawfulness of content); Force v. Facebook, Inc., 934 F.3d 53, 66–69 (2d Cir. 2019) (rejecting theories that algorithmic sorting rendered website a nonpublisher or materially contributed to development of content); Marshall’s Locksmith Serv., Inc. v. Google, LLC, 925 F.3d 1263, 1271 (D.C. Cir. 2019) (declining to treat search engines’ conversion of fraudulent addresses from webpages into “map pinpoints” as developing content). 179 Force v. Facebook, Inc., 934 F.3d 53, 57 (2d Cir. 2019). 180 Id. at 70 (internal quotations omitted); id. at 65–66. 181 Id. at 71. In a partially dissenting opinion, Chief Judge Katzmann wrote that he would not apply Section 230(c)(1), reasoning that claims based on Facebook’s friend and content suggestion systems did not treat Facebook as a publisher of another’s content. Id. at 76–89 (Katzmann, J., concurring in part and dissenting in part). 182 Id. at 66–67 (majority opinion); see Carafano v. Metrosplash.com, Inc., 339 F.3d 1119, 1124–25 (9th Cir. 2003) (applying Section 230 to a website’s “decision to structure the information provided by users”); Marshall’s Locksmith Serv., 925 F.3d at 1269 (holding that “the choice of presentation” is a publisher function protected by Section 230); cf. O’Kroley v. Fastcase, Inc., 831 F.3d 352, 355 (6th Cir. 2016) (applying Section 230 to “automated editorial acts”). 183 Force, 934 F.3d at 70. 184 Id. at 66 (“[W]e find no basis . . . for concluding that an interactive computer service is not the ‘publisher’ of third- party information when it uses tools such as algorithms that are designed to match that information with a consumer’s interests.”) (citing Fair Hous. Council v. Roommates.com, Inc., 521 F.3d 1157, 1172 (9th Cir. 2008)).

Congressional Research Service 18 Section 230: An Overview

does not create or develop content when it merely provides a neutral means by which third parties can post information of their own independent choosing online.”185 Both the D.C. Circuit and the Second Circuit have elaborated on particular features that may make a website’s tools “neutral.” In Marshall’s Locksmith Service v. Google, a case involving search engines that automatically converted addresses provided by third parties into “pinpoints” appearing on the search engines’ mapping websites, the D.C. Circuit emphasized that the search engines’ tools did “not distinguish” between different types of user content.186 Instead, the algorithm translated all types of information, both legitimate and scam information, in the same manner.187 The Second Circuit in Force characterized Facebook’s involvement in user content as “neutral” when Facebook did not require users to provide more than “basic identifying information” and its sorting algorithms used “objective factors” that applied in the same way “to any content.”188

Section 230(c)(2)(A): Restricting Access to Objectionable Material Section 230(c)(2)(A) states that service providers and users may not “be held liable” for voluntary, “good faith” actions “to restrict access to or availability of material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally protected.”189 This provision is more limited than Section 230(c)(1) in a few ways. First, as discussed above,190 while a number of courts have held that Section 230(c)(1) shields decisions both to distribute and to restrict others’ content, Section 230(c)(2) applies only to decisions to restrict content. For example, providers have successfully invoked Section 230(c)(2) in claims challenging decisions to remove user videos,191 suspend accounts,192 prevent unsolicited bulk emails,193 or decisions not to run certain ads.194 In addition, unlike Section 230(c)(1), Section 230(c)(2) applies only to voluntary, good-faith actions, and it applies only to the listed categories of “objectionable” material.195

185 Klayman v. Zuckerberg, 753 F.3d 1354, 1358 (D.C. Cir. 2014); accord Kimzey v. Yelp! Inc., 836 F.3d 1263, 1270 (9th Cir. 2016) (characterizing a rating system based on third-party input as a “neutral tool”). 186 Marshall’s Locksmith Serv., 925 F.3d at 1271. 187 Id. 188 Force, 934 F.3d at 70. 189 47 U.S.C. § 230(c)(2)(A). 190 Supra note 76 and accompanying text. 191 Domen v. Vimeo, Inc., 433 F. Supp. 3d 592, 603–04 (S.D.N.Y. 2020) (involving free speech and discrimination claims). 192 Dipp-Paz v. Facebook, No. 18-CV-9037, 2019 WL 3205842, at *3 (S.D.N.Y. July 12, 2019) (involving constitutional free speech claims); Cox v. Twitter, Inc., No. 2:18-2573-DCN-BM, 2019 WL 2513963, at *3 (D.S.C. Feb. 8, 2019) (involving unclear legal claims). 193 Green v. Am. Online (AOL), 318 F.3d 465, 473 (3rd Cir. 2003) (involving negligence, breach of contract, constitutional free speech, and consumer fraud claims); Holomaxx Techs. v. Microsoft Corp., 783 F. Supp. 2d 1097, 1105 (N.D. Cal. 2011) (involving, among others, intentional interference with contract and intentional interference with prospective business advantage claims); 360Insight, LLC v. Comcast Corp., 546 F. Supp. 2d 605, 607 (N.D. Ill. 2008) (involving federal Computer Fraud and Abuse Act, constitutional free speech, tortious interference with prospective economic advantage, and consumer fraud claims). 194 Langdon v. Google, Inc., 474 F. Supp. 2d 622, 630–31 (D. Del. 2007) (involving free speech, fraud, breach of contract, deceptive business practices, and “public calling” claims). 195 See 47 U.S.C. § 230(c). See also, e.g., Fyk v. Facebook, Inc., 808 Fed. App’x 597, 598 (9th Cir. 2020) (“Unlike 47 U.S.C. § 230(c)(2)(A), nothing in § 230(c)(1) turns on the alleged motives underlying the editorial decisions of the provider of an interactive computer service.”).

Congressional Research Service 19 Section 230: An Overview

These limits on Section 230(c)(2) immunity have been litigated in the courts and have led courts to conclude, in some circumstances, that providers cannot claim Section 230 immunity.196

Good Faith Providers or users may claim immunity under Section 230(c)(2)(A) only if they act in “good faith.”197 The statute does not itself define what it means to act in good faith, and courts have applied a few different understandings of the term. Some trial court decisions have denied immunity and allowed claims to proceed where the plaintiff alleged that a service provider acted with an anticompetitive motive.198 For example, one court declined to dismiss a lawsuit alleging that Google had engaged in unfair competition by removing a company’s websites from its search results.199 Although Google said it had removed the results because they were “webspam” that violated its guidelines, the plaintiff claimed that Google actually had acted with an anticompetitive motive, because the plaintiff, which specialized in search engine optimization, “was cutting into Google’s revenues.”200 The court ruled that the plaintiff had presented enough evidence “to raise a genuine issue of fact” as to whether Google acted in good faith, preventing the court from dismissing the claim under Section 230.201 To take another example, a different court allowed a claim to proceed where the plaintiff alleged that YouTube removed her video to punish her for working with a competitor rather than buying Google’s advertising services.202 In evaluating whether a provider acted in good faith, courts have also looked to whether the provider’s rationale for restricting content is “pretextual.”203 As one trial court put it, for a removal to be made in good faith, “the provider must actually believe that the material is objectionable for the reasons it gives.”204 Under this view, if a provider says it is enforcing its terms of service, but is in fact motivated by some other reason, the provider may be acting in bad faith.205 Another court concluded that a service provider could be seen as acting in bad faith when

196 See, e.g., Enhanced Athlete Inc. v. Google LLC, No. 19-cv-08260-HSG, 2020 WL 4732209, at *4 (N.D. Cal. Aug. 14, 2020); e-ventures Worldwide, LLC v. Google, Inc., No. 2:14-cv-646-FtM-PAM-CM, 2017 WL 2210029, at *3 (M.D. Fla. Feb. 8, 2017); Darnaa, LLC v. Google, Inc., No. 15-cv-03221-RMW, 2016 WL 6540452, at *8 (N.D. Cal. Nov. 2, 2016). 197 47 U.S.C. § 230(c)(2)(A). 198 See Darnaa, 2016 WL 6540452, at *8–9 (involving allegation that Google removed plaintiff’s video from YouTube because the plaintiff refused to allow Google to embed advertising in the video); e-ventures Worldwide, LLC, 2017 WL 2210029, at *1, 3 (M.D. Fla. Feb. 8, 2017) (involving allegation that Google delisted plaintiff, “an online publisher that specializes in” search engine optimization, because it was “cutting into Google’s revenues”). Cf. Spy Phone Labs LLC v. Google Inc., No. 15-cv-03756-KAW, 2016 WL 6025469, at *8 (N.D. Cal. Oct. 14, 2016) (involving allegation that Google was retaliating against plaintiff for submitting a trademark infringement complaint against another app). 199 e-ventures Worldwide, LLC, 2017 WL 2210029, at *1–2. Specifically, the lawsuit involved claims of “unfair competition under the Lanham Act, 15 U.S.C. § 1125(a); violation of Florida’s Deceptive and Unfair Trade Practices Act; and tortious interference with contractual relationships.” Id. at *2. 200 Id. at *1. 201 Id. at *3. 202 Darnaa, 2016 WL 6540452, at *8–9. 203 Spy Phone Labs LLC v. Google Inc., No. 15-cv-03756-KAW, 2016 WL 6025469, at *8 (N.D. Cal. Oct. 14, 2016); accord GCM Partners, LLC v. Hipaaline Ltd., No. No. 20 C 6401, 2020 WL 6867207, at *13 (N.D. Ill. Nov. 23, 2020), appeal docketed, No. 20-3509 (7th Cir. Dec. 23, 2020). 204 Darnaa, 2016 WL 6540452, at *8. 205 Id.; Spy Phone Labs, 2016 WL 6025469, at *8. But see Langdon v. Google, Inc., 474 F. Supp. 2d 622, 631 (D. Del. 2007) (rejecting plaintiff’s assertion that the provider did not act in good faith because it gave false reasons for declining to run his ads, on the grounds that the provider must have permissibly concluded they were “otherwise objectionable”).

Congressional Research Service 20 Section 230: An Overview

the provider “failed to respond to [the user’s] repeated requests for an explanation.”206 In comparison, one trial court suggested that “selective enforcement” of a policy would not be enough, by itself, to demonstrate bad faith.207

Objectionable Material The other important limitation on Section 230(c)(2)(A) immunity is that it applies only when providers or users restrict the listed types of content: “material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable.”208 Although this list includes only specific types of content, it can still be interpreted relatively broadly. In particular, some courts have interpreted the catch-all phrase “otherwise objectionable” broadly because Section 230(c)(2)(A) states that the provider or user is the one who determines whether the content is objectionable.209 As one court noted, the statute’s text injects “a subjective element” into this inquiry, by asking whether “the provider or user considers” the content to be objectionable.210 Thus, some courts have concluded, without significant discussion, that material classified as spam or malware can be considered “harassing” or “objectionable” under Section 230(c)(2)(A).211 In some cases, courts have looked to providers’ policies to determine whether the providers considered the restricted material objectionable.212 In 2009, one Ninth Circuit judge expressed concern about interpreting “otherwise objectionable” too broadly, cautioning that “the literal terms of” Section 230(c)(2)(A) could be read to grant providers “free license to unilaterally block the dissemination of material by content providers.”213 While the “good faith” provision discussed above limits providers’ discretion,214 some courts have concluded that “otherwise objectionable” should also be read more narrowly to avoid giving providers this free license.215 For example, one trial court denied Section 230

206 Smith v. Trusted Universal Standards in Elec. Transactions, Inc., No. 09-4567, 2011 WL 900096, at *9 (D.N.J. Mar. 15, 2011). 207 Spy Phone Labs, 2016 WL 6025469, at *8. See also e360Insight, LLC v. Comcast Corp., 546 F. Supp. 2d 605, 609 (N.D. Ill. 2008) (ruling that plaintiff did not sufficiently plead an “absence of good faith” even though the plaintiff claimed the provider “singl[ed] out” the plaintiff). 208 47 U.S.C. § 230(c)(2)(A). 209 See, e.g., e360Insight, 546 F. Supp. 2d at 607–08. 210 Id. at 608. See also Domen v. Vimeo, Inc., 433 F. Supp. 3d 592, 603 (S.D.N.Y. 2020) (“Section 230(c)(2) is focused upon the provider’s subjective intent . . . .”); Zango, Inc. v Kaspersky Lab, Inc., No. 07-0807-JCC, 2007 WL 5189857, at *4 (W.D. Wash. Aug. 28, 2007) (“Section 230(c)(2)(A), which provides the definition of the relevant material described in Section 230(c)(2)(B), does not require that the material actually be objectionable; rather, it affords protection for blocking material ‘that the provider or user considers to be’ objectionable. 47 U.S.C. § 230(c)(2)(A).”), aff’d, 568 F.3d 1169 (9th Cir. 2009). Cf. Holomaxx Techs. v. Microsoft Corp., 783 F. Supp. 2d 1097, 1104 (N.D. Cal. 2011) (“No court has articulated specific, objective criteria to be used in assessing . . . a provider’s subjective determination of what is ‘objectionable’ . . . . Here, however, it is clear . . . that Microsoft reasonably could conclude that Holomaxx’s emails were ‘harassing’ and thus ‘otherwise objectionable.’” (emphasis added)). 211 e-ventures Worldwide, LLC v. Google, Inc., No. 2:14-cv-646-FtM-PAM-CM, 2017 WL 2210029, at *3 (M.D. Fla. Feb. 8, 2017) (“[S]pam is undoubtedly ‘harassing’ or ‘objectionable’ content for purposes of the CDA.”); accord Green v. Am. Online (AOL), 318 F.3d 465, 473 (3rd Cir. 2003). See also Zango, 2007 WL 5189857, at *4 (“There is no question that [the provider] considers the software to be objectionable [as malware].”); Langdon v. Google, Inc., 474 F. Supp. 2d 622, 631 (D. Del. 2007) (concluding implicitly, without discussion, that Section 230 barred plaintiff’s lawsuit because Google considered his ads “otherwise objectionable”). 212 E.g., Domen, 433 F. Supp. 3d at 604; e360Insight, 546 F. Supp. 2d at 608. 213 Zango, Inc. v. Kaspersky Lab, Inc., 568 F.3d 1169, 1178 (9th Cir. 2009) (Fisher, J., concurring). 214 Cf. id. at 1179 (expressing concern that Section 230(c)(2)(B) does not contain a good faith limitation). 215 See, e.g., Enigma Software Grp. USA, LLC v. Malwarebytes, Inc., 946 F.3d 1040, 1050 (9th Cir. 2019); Song Fi

Congressional Research Service 21 Section 230: An Overview

immunity to YouTube in a case challenging YouTube’s decision to remove a video because its view count had allegedly been artificially inflated.216 The court noted that the ordinary meaning of “objectionable” could include anything a provider finds undesirable, but ultimately concluded that such a broad definition was inconsistent with “the context, history, and purpose” of Section 230.217 Looking to the list of adjectives preceding “otherwise objectionable,” the court believed that Congress was focused on “potentially offensive materials, not simply any materials undesirable to a content provider or user.”218 Consequently, the court said that “it is hard to imagine that the phrase includes . . . the allegedly artificially inflated view count.”219 Similarly looking to congressional intent, the Ninth Circuit held in a 2019 case that the term “otherwise objectionable” should be interpreted to exclude anticompetitive conduct.220 This Ninth Circuit ruling interpreted Section 230(c)(2)(B) and is discussed in more detail below.221

Section 230(c)(2)(B): Enabling Access Restriction Section 230(c)(2)(B) provides that service providers and users may not “be held liable” for actions “taken to enable or make available to . . . others the technical means to restrict access to material” that falls within the specific categories listed in Section 230(c)(2)(A).222 Accordingly, Section 230(c)(2)(B) focuses on enabling others to restrict access to objectionable material, and offers immunity to, for example, “providers of programs that filter adware and malware,”223 as well as services that enable the filtering of spam email.224 Courts have concluded that companies like Facebook are also eligible for Section 230(c)(2)(B) immunity, to the extent they provide users with tools to hide or otherwise restrict their own access to content.225

Inc. v. Google, Inc., 108 F. Supp. 3d 876, 884 (N.D. Cal. 2015); Darnaa, LLC v. Google, Inc., No. 15-cv-03221-RMW, 2016 WL 6540452, at *8 (N.D. Cal. Nov. 2, 2016). 216 Song Fi, 108 F. Supp. 3d at 882. 217 Id. at 882, 884. 218 Id. See also Darnaa, 2016 WL 6540452, at *8 (“The context of § 230(c)(2) appears to limit the term [“objectionable”] to that which the provider or user considers sexually offensive, violent, or harassing in content.”); but see Enigma Software Grp. USA, 946 F.3d at 1051 (rejecting the argument that “the term ‘objectionable’ . . . cover[s] only material that is sexual or violent in nature”). 219 Song Fi, 108 F. Supp. 3d at 883. 220 Enigma Software Grp. USA, 946 F.3d at 1045 (“We hold that the phrase ‘otherwise objectionable’ does not include software that the provider finds objectionable for anticompetitive reasons.”); id. at 1051 (“Congress wanted to encourage the development of filtration technologies, not to enable software developers to drive each other out of business.”). 221 Infra notes 236 to 242 and accompanying text. 222 47 U.S.C. § 230(c)(2)(B). Although Section 230(c)(2)(B) refers to “material described in paragraph (1),” a note in the United States Code indicates that this is likely meant to reference “subparagraph (A)” instead. Id. n.1. 223 Zango, Inc. v. Kaspersky Lab, Inc., 568 F.3d 1169, 1174 (9th Cir. 2009). See generally, e.g., Russell A. Miller, The Legal Fate of Internet Ad-Blocking, 24 B.U. J. SCI. & TECH. L. 301, 358–60 (2018) (discussing how Section 230(c)(2)(B) might protect ad-blocking firms from liability). 224 Smith v. Trusted Universal Standards in Elec. Transactions, Inc., No. 09-4567, 2011 WL 900096, at *6 (D.N.J. Mar. 15, 2011) (granting Section 230(c)(2)(B) immunity to service that investigated and provided information about IP addresses, “help[ing] information content providers restrict access to spam email”); id. at *8 (granting Section 230(c)(2)(B) immunity to software that “provide[d] Comcast with a means to restrict access to harassing spam email”). 225 Fehrenbach v. Zeldin, No. 17-CV-5282, 2018 WL 4242452, at *5 (E.D.N.Y. Aug. 6, 2018) (holding that Section 230(c)(2)(B) immunized Facebook from a complaint premised on the fact that Facebook allows users to hide comments).

Congressional Research Service 22 Section 230: An Overview

The fact that a company provides users with the choice to opt out of receiving certain content, however, may not always be sufficient to gain Section 230(c)(2)(B) immunity.226 In one case, a plaintiff sued Yahoo! for sending automated text message notifications about messages the plaintiff had received on Yahoo! Messenger.227 Yahoo! claimed that the suit was barred by Section 230(c)(2)(B) because the text message “include[d] a link to a help page which . . . contain[ed] instructions on how to block further messages,” and accordingly, made “available the ‘technical means to restrict access’ to messages which plaintiff might deem ‘objectionable.’”228 The trial court rejected this claim, noting that because the text message notifications were sent automatically, “neither Yahoo! nor the mobile phone user ha[d] the opportunity to determine whether the third party message” was objectionable.229 Accordingly, the court held that Yahoo! could not claim Section 230(c)(2)(B) immunity where it “did not engage in any form of content analysis of the subject text to identify material that was offensive or harmful prior to the automatic sending of a notification message.”230 Because Section 230(c)(2)(B) applies only to actions restricting the types of content listed in Section 230(c)(2)(A),231 it implicates the same interpretive questions discussed above regarding whether the provider or user considered the restricted material “to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable.”232 However, unlike Section 230(c)(2)(A), Section 230(c)(2)(B) does not contain an explicit requirement for the provider or user to act in good faith.233 Thus, one Ninth Circuit judge expressed concern that Section 230(c)(2)(B) could be read to grant immunity to bad faith conduct, including “covert, anticompetitive blocking” of competitors.234 The judge believed it was “very likely” that Congress “did not intend to immunize” such conduct.235 In a 2019 decision, Enigma Software Group USA, LLC v. Malwarebytes, Inc., the Ninth Circuit held that Section 230(c)(2)(B) did not block a suit alleging anticompetitive conduct.236 A company that sold computer security software sued a competitor after the competitor began flagging some of the plaintiff’s programs as “potentially unwanted programs.”237 The plaintiff argued that this characterization served “as a ‘guise’ for anticompetitive conduct.”238 In evaluating

226 Sherman v. Yahoo! Inc., 997 F. Supp. 2d 1129, 1138 (S.D. Cal. 2014). 227 Id. at 1130. 228 Id. at 1137 (quoting 47 U.S.C. § 230(c)(2)). 229 Id. at 1138. 230 Id. 231 47 U.S.C. § 230(c)(2)(B). 232 Id. § 230(c)(2)(A); supra “Objectionable Material.” 233 47 U.S.C. § 230(c)(2). See also, e.g., Zango, Inc. v. Kaspersky Lab, Inc., 568 F.3d 1169, 1177 (9th Cir. 2009) (holding that allegations that provider acted in bad faith did not preclude dismissal of suit under Section 230(c)(2)(B) because this subparagraph “has no good faith language,” and noting that the plaintiff waived any argument that the provision “should be construed implicitly to have a good faith component”). 234 Zango, 568 F.3d at 1179 (Fisher, J., concurring). 235 Id.; see also id. at 1179 n.3 (“[T]he legislative history the parties cite is not helpful in determining the exact boundaries of what Congress intended to immunize. Whatever those exact boundaries, I doubt Congress intended to leave victims of malicious or anticompetitive blocking without a cause of action . . . .”). 236 Enigma Software Grp. USA, LLC v. Malwarebytes, Inc., 946 F.3d 1040, 1045 (9th Cir. 2019). 237 Id. at 1047–48. 238 Id. at 1048. Specifically, the complaint alleged both state law causes of action—deceptive business practices and tortious interference with business and contractual relations—and a federal claim under the Lanham Act. Id. The Ninth Circuit also considered whether the Lanham Act claim fell within the Section 230 exception for intellectual property claims, holding that it did not. Id. at 1045.

Congressional Research Service 23 Section 230: An Overview

the competitor’s attempt to claim immunity under Section 230(c)(2)(B), the Ninth Circuit looked to Congress’s purpose in enacting Section 230, concluding that “Congress wanted to encourage the development of filtration technologies, not to enable software developers to drive each other out of business.”239 Accordingly, the court rejected the idea that the competitor could claim immunity “regardless of anticompetitive purpose.”240 The court believed that the term “objectionable” is not limited only to “material that is sexual or violent in nature,” and can encompass other “forms of unwanted online content that Congress could not identify in the 1990s.”241 But “if a provider’s basis for objecting to and seeking to block materials is because those materials benefit a competitor,” the court held that the provider could not claim Section 230 immunity.242 This decision was appealed to the Supreme Court, and although the Court declined the appeal, the case garnered a number of amicus briefs from parties interested in the case, as well as a separate statement from Justice Thomas respecting the denial of certiorari.243 Interest groups argued that the Ninth Circuit’s decision improperly imported a “good faith” requirement into Section 230(c)(2)(B), even though the text did not contain such a limitation.244 In an opinion concurring in the Court’s decision to deny certiorari, Justice Thomas argued that the Ninth Circuit decision— and other decisions interpreting Section 230—improperly “relied on purpose and policy” rather than textual arguments, creating “questionable precedent.”245 It remains to be seen whether courts outside the Ninth Circuit will agree with its ruling.

Section 230(e): Exceptions As detailed above, Section 230(e) outlines five exceptions to the immunity created by Section 230.246 A defendant cannot claim Section 230 immunity as a basis to dismiss a federal criminal prosecution or any lawsuit brought under intellectual property laws, state laws that are consistent with Section 230, certain electronic communications privacy laws, or certain sex trafficking laws.247

Federal Criminal Law The first exception to Section 230 immunity is for “any . . . Federal criminal statute,” meaning that any defendant in a federal criminal prosecution cannot claim Section 230 immunity.248 For example, Section 230 does not bar prosecution under federal statutes that prohibit the knowing

239 Id. at 1051. 240 Id. 241 Id. at 1051–52. 242 Id. at 1052. However, the court noted that the defendant provider disputed whether it did engage in “anticompetitive blocking” and claimed instead that it found the plaintiff’s “programs ‘objectionable’ for legitimate reasons based on the programs’ content.” Id. The court suggested this factual dispute could be resolved on remand to the lower court. Id. 243 See Malwarebytes, Inc. v. Enigma Software Grp. USA, LLC, 141 S. Ct. 13 (2020). 244 See, e.g., Brief of Electronic Frontier Foundation as Amicus Curiae in Support of Petitioner at 4, Malwarebytes, Inc., 208 L. Ed. 2d 197 (No. 19-1284); Brief of TechFreedom as Amicus Curiae in Support of Petitioner at 5, Malwarebytes, Inc., 208 L. Ed. 2d 197 (No. 19-1284). 245 Malwarebytes, Inc., 141 S. Ct. at 13–14 (Thomas, J., statement respecting the denial of certiorari). 246 Supra notes 35 to 44 and accompanying text. 247 See 47 U.S.C. § 230(e). 248 Id. § 230(e)(1).

Congressional Research Service 24 Section 230: An Overview

distribution of obscene materials.249 Neither did Section 230 bar the federal prosecution of Backpage.com corporate entities for conspiracy to engage in money laundering.250 This exception does not include state criminal laws, and courts have read Section 230 to preempt inconsistent state criminal prosecutions.251 Courts have interpreted Section 230(e)(1) to allow only criminal prosecutions, not civil lawsuits based on violations of federal criminal laws.252 A number of plaintiffs have argued that, particularly where federal law creates criminal and civil liability for the same conduct, applying Section 230 to bar suits under a civil enforcement provision would “impair the enforcement” of the criminal law.253 Courts have rejected those arguments,254 noting the traditional distinction between criminal and civil liability and concluding that, by referring only to “criminal” statutes in Section 230(e)(1), Congress intended to exclude civil suits.255

Intellectual Property Law The second exception to Section 230 immunity is for “any law pertaining to intellectual property.”256 This phrase is somewhat ambiguous,257 but courts have concluded that this exception for laws “pertaining to intellectual property” allows, for example, suits for copyright and trademark infringement.258 In evaluating whether Section 230(e)(2) applies, courts have sometimes looked not only to whether the plaintiff is suing under a law that generally involves

249 See, e.g., 18 U.S.C. § 1462 (making it a crime to “knowingly use[] any . . . interactive computer service . . . for carriage in interstate or foreign commerce—(a) any obscene, lewd, lascivious, or filthy . . . picture, motion-picture film, . . . writing, print, or other matter of indecent character; or (b) any obscene, lewd, lascivious, or filthy . . . electrical transcription, or other article or thing capable of producing sound”). 250 See Press Release, U.S. Dep’t of Justice, Backpage’s Co-founder and CEO, As Well As Several Backpage-Related Corporate Entities, Enter Guilty Pleas (Apr. 12, 2018), https://www.justice.gov/opa/pr/backpage-s-co-founder-and-ceo- well-several-backpage-related-corporate-entities-enter-guilty. 251 See generally, e.g., Voicenet Commc’ns, Inc. v. Corbett, No. 04-1318, 2006 WL 2506318, at *3 (E.D. Pa. Aug. 30, 2006) (interpreting Section 230(e)(1) not to include state criminal laws); see also, e.g., Universal Commc’n Sys., Inc. v. Lycos, Inc., 478 F.3d 413, 422 (1st Cir. 2007) (dismissing suit under state cyberstalking law because defendant’s “liability would depend on treating it as the publisher of those postings”); Backpage.com, LLC v. McKenna, 881 F. Supp. 2d 1262, 1273 (W.D. Wash. 2012) (concluding proposed state legislation “is likely inconsistent with and therefore expressly preempted by [47 U.S.C. § 230]” because it imposes “liability on Backpage.com and [Internet Archive] for information created by third parties—namely ads for commercial sex acts depicting minors—so long as it ‘knows’ that it is publishing, disseminating, displaying . . . such information”). 252 See, e.g., Gonzalez v. Google, Inc., 335 F. Supp. 3d 1156, 1169 (N.D. Cal. 2018). Other exceptions do allow specific federal civil claims; for example, civil suits based on certain federal sex trafficking offenses may be permitted under a different exception. See infra “Sex Trafficking Law (FOSTA).” 253 E.g., Force v. Facebook, Inc., 934 F.3d 53, 71 (2d Cir. 2019); Doe v. Backpage.com, LLC, 817 F.3d 12, 23 (1st Cir. 2016); Doe v. Bates, No. 5:05CV91, 2006 WL 8440858, at *13 (E.D. Tex. Jan. 18, 2006). 254 E.g., Force, 934 F.3d at 72; Backpage.com, 817 F.3d at 23; Bates, 2006 WL 8440858, at *14. But see Nieman v. Versuslaw, Inc., No. 12-3104, 2012 WL 3201931, at *9 (C.D. Ill. Aug. 3, 2012) (saying in dicta that “arguably, § 230 of the CDA may not be used to bar a civil RICO claim because that would impair the enforcement of a Federal criminal statute”). 255 See, e.g., Force, 934 F.3d at 71; Backpage.com, 817 F.3d at 23. 256 47 U.S.C. § 230(e)(2). 257 See Perfect 10, Inc. v. CCBill LLC, 488 F.3d 1102, 1119 (9th Cir. 2007) (“The CDA does not contain an express definition of ‘intellectual property,’ and there are many types of claims in both state and federal law which may—or may not—be characterized as ‘intellectual property’ claims.”). 258 E.g., Parker v. Google, Inc., 422 F. Supp. 2d 492, 503 n.8 (E.D. Penn. 2006); Gucci Am., Inc. v. Hall & Assocs., 135 F. Supp. 2d 409, 414 (S.D.N.Y. 2001); Malibu Media, LLC v. Weaver, No. 8:14-cv-1580-T-33TBM, 2016 WL 1394331, at *8 (M.D. Fla. Apr. 8, 2016).

Congressional Research Service 25 Section 230: An Overview

intellectual property issues, but more specifically, whether the plaintiff’s claim itself involves an intellectual property right.259 For example, the Ninth Circuit ruled in 2019 that a false advertising claim brought under the Lanham Act did not fall within the Section 230(e)(2) exception.260 The court noted that the Lanham Act, a federal law, “contains two parts, one governing trademark infringement and another governing false designations of origin, false descriptions, and dilution.”261 Noting that Congress intended to provide broad immunity in Section 230, the Ninth Circuit construed the intellectual property exception narrowly, to include only “claims pertaining to an established intellectual property right . . . like those inherent in a patent, copyright, or trademark.”262 Because the false advertising claim did not “relate to or involve trademark rights or any other intellectual property rights,” the court held that the intellectual property exception did not apply.263 Somewhat similarly, a New Hampshire trial court held in one case that three “right of privacy” torts— intrusion upon seclusion, publication of private facts, and casting in false light—involved rights that could not be considered property rights.264 Accordingly, the court concluded that the claims did “not sound in ‘law pertaining to intellectual property’” and Section 230 barred the claims.265 Courts have disagreed about whether Section 230(e)(2) includes state law claims such as the right to publicity,266 a cause of action that essentially allows plaintiffs to sue for the improper commercial use of their identity.267 Some courts have held that the exception does include state intellectual property claims, allowing, for example, state law claims for copyright infringement, misappropriation and unfair competition, and right of publicity to proceed.268 These courts have noted that the exception refers broadly to “any law,”269 and that other provisions of Section 230

259 Enigma Software Grp. USA, LLC v. Malwarebytes, Inc., 946 F.3d 1040, 1052–53 (9th Cir. 2019). See also, e.g., Corker v. Costco Wholesale Corp., No. C19-0290RSL, 2019 WL 5895430, at *6 (W.D. Wash. Nov. 12, 2019) (concluding Section 230(e)(2) did not apply to a false association claim because the claim did “not involve an intellectual property right or trademark”); Doe v. Friendfinder Network, Inc., 540 F. Supp. 2d 288, 302–03 (D.N.H. 2008) (holding that Section 230(e)(2) did not apply to state right of privacy claims that involved personal rights). 260 Enigma Software Grp. USA, 946 F.3d at 1053. However, the court nonetheless concluded that because the claim was “based on allegations of [anticompetitive] conduct,” it would not apply Section 230 to dismiss the claim. Id. at 1054. 261 Id. at 1053. 262 Id. 263 Id. at 1053–54. 264 Friendfinder Network, Inc., 540 F. Supp. 2d at 302–03. 265 Id. at 303. 266 See, e.g., Stayart v. Yahoo! Inc., 651 F. Supp. 2d 873, 888–89 (W.D. Wis. 2009) (noting that a right to publicity claim “is generally considered an intellectual property claim,” implicating this exception, but further noting the “disagreement among various federal courts regarding the scope of the intellectual property exception,” and ultimately dismissing the claim on jurisdictional grounds); see also Friendfinder Network, Inc., 540 F. Supp. 2d at 302 (holding that a state right of publicity claim “arises out of a ‘law pertaining to intellectual property’ within the meaning of” 47 U.S.C. § 230(e)(2)). 267 See, e.g., ETW Corp. v. Jireh Publ’g, Inc., 332 F.3d 915, 928–37 (6th Cir. 2003) (discussing the right of publicity). 268 Atl. Recording Corp. v. Project Playlist, Inc., 603 F. Supp. 2d. 690, 694, 704 (S.D.N.Y. 2009); Friendfinder Network, Inc., 540 F. Supp. 2d at 302. The First Circuit suggested in one decision that a state trademark claim was “not subject to Section 230 immunity.” Universal Commc’n Sys., Inc. v. Lycos, Inc., 478 F.3d 413, 422–23 (1st Cir. 2007). However, this conclusion was arguably dicta, given that the First Circuit ultimately concluded that the claim “would not survive” even if Section 230 did not apply. Id. at 423; see also Friendfinder Network, Inc., 540 F. Supp. 2d at 299 (describing this statement as dicta). 269 47 U.S.C. § 230(e)(2) (emphasis added).

Congressional Research Service 26 Section 230: An Overview

distinguish between state and federal law, suggesting that “any law” includes both state and federal laws.270 Other courts, including the Ninth Circuit, have held that Section 230(e)(2) encompasses only federal laws and that Section 230 bars state intellectual property claims.271 In Perfect 10, Inc. v. CCBill LLC, the Ninth Circuit looked to Congress’s policy goals and “construe[d] the term ‘intellectual property’ to mean ‘federal intellectual property.’”272 The court noted that state intellectual property laws “are by no means uniform,” and could subject websites to varied liability schemes.273 In the view of the court, this outcome “would be contrary to Congress’s expressed goal of insulating the development of the Internet from the various state-law regimes.”274 The Ninth Circuit did not discuss the fact that the text of Section 230(e)(2) refers to “any law,” noting only that the term “intellectual property” was undefined.275

State Law The third exception provides that Section 230 will not “prevent any State from enforcing any State law that is consistent with this section,” allowing states to continue enforcing any laws that are “consistent” with Section 230.276 As one trial court described this provision, “Section 230(e)(3) does not attempt to define what state law is consistent and inconsistent with the CDA;” in effect, this subsection “provides no substantive content.”277 In evaluating whether a state law, or a particular application of a state law, is consistent with Section 230 or whether it is instead inconsistent and preempted, courts have looked to whether the law would violate Section 230(c)(1) by treating service providers or users as the publisher of another person’s content.278 Accordingly, for example, one court concluded that a libel claim that would hold a website operator “liable for statements he actually authored” was “consistent with” Section 230 and could proceed.279

270 Atl. Recording Corp., 603 F. Supp. 2d. at 703–04; Friendfinder Network, Inc., 540 F. Supp. 2d at 299–300. 271 See, e.g., Perfect 10, Inc. v. CCBill LLC, 488 F.3d 1102, 1119 (9th Cir. 2007); Hepp v. Facebook, Inc., 465 F. Supp. 3d 491, 500 (E.D. Penn. 2020). 272 Perfect 10, Inc., 488 F.3d at 1118–19. 273 Id. at 1118. 274 Id. See also Hepp, 465 F. Supp. 3d at 500 (“This Court is persuaded by the reasoning in Perfect 10. . . . Conditioning CDA immunity on the diverse potentially applicable state laws . . . would run contrary to the purpose and intent of the CDA.”). 275 See Perfect 10, Inc., 488 F.3d at 1119; see also, e.g., Friendfinder Network, Inc., 540 F. Supp. 2d at 299 (“The Ninth Circuit made no attempt to reckon with the presence of the term ‘any’—or for that matter, the absence of term ‘federal’—in § 230(e)(2) when limiting it to federal intellectual property laws.”). 276 47 U.S.C. § 230(e)(3). 277 Atl. Recording Corp. v. Project Playlist, Inc., 603 F. Supp. 2d. 690, 694, 702 (S.D.N.Y. 2009). 278 Compare, e.g., HomeAway.com, Inc. v. City of Santa Monica, 918 F.3d 676, 683 (2019) (holding that an ordinance regulating home rentals “is not ‘inconsistent’ with the CDA” because it would not impose a duty on websites to monitor third-party content), with, e.g., Backpage.com, LLC v. McKenna, 881 F. Supp. 2d 1262, 1273 (W.D. Wash. 2012) (holding that a state criminal law “is likely inconsistent with and therefore expressly preempted by Section 230” because it would impose liability on websites for third-party content). 279 Cisneros v. Sanchez, 403 F. Supp. 2d 588, 592 (S.D. Tex. 2005) (emphasis added).

Congressional Research Service 27 Section 230: An Overview

Electronic Communications Privacy Act of 1986 The fourth exception to Section 230 immunity is for claims brought under the Electronic Communications Privacy Act of 1986 (ECPA) “or any similar State law.”280 ECPA is a federal law that governs wiretapping and electronic eavesdropping.281 ECPA creates a number of criminal offenses, which would fall within the first exception for federal crimes,282 but also contains civil liability provisions, which fall within this fourth exception.283 Perhaps most relevant to service providers that host user-generated content, ECPA makes it unlawful not only to intercept covered communications intentionally, but also to disclose information intentionally if the person “ha[s] reason to know that the information was obtained through” an unlawful interception.284 However, the Seventh Circuit ruled in one case that this exception did not allow a lawsuit against companies that provided web hosting services to people who sold illegally obtained videos.285 The court said the plaintiffs had not shown that the web service companies had “disclose[d] any communication,” declining to impose secondary liability on the web service providers absent evidence that the companies provided “culpable assistance” to the “wrongdoer.”286

Sex Trafficking Law (FOSTA) The fifth exception to Section 230 immunity was added in 2018 by the Allow States and Victims to Fight Online Sex Trafficking Act of 2017 (FOSTA), and creates exceptions for certain sex trafficking offenses.287 Specifically, Section 230(e)(5) provides that Section 230 will not bar: (1) federal civil actions288 “if the conduct underlying the charge would constitute a violation of” 18 U.S.C. § 1591, which prohibits knowingly engaging in sex trafficking of minors, or in sex trafficking that involves force, fraud, or coercion; (2) state criminal prosecutions where the underlying conduct would violate 18 U.S.C. § 1591; and (3) state criminal prosecutions where the underlying conduct would violate 18 U.S.C. § 2421A, which prohibits “operat[ing] an interactive computer service . . . with the intent to promote or facilitate the prostitution of another person” in jurisdictions where such conduct is illegal.289 The FOSTA exceptions will apply only if a private plaintiff or state prosecutor can demonstrate that the service provider or user violated the specified federal laws.290 Accordingly, one federal trial court interpreting the first FOSTA exception, for federal civil lawsuits, concluded that the

280 47 U.S.C. § 230(e)(4). See also, e.g., Universal Commc’n Sys., Inc. v. Lycos, Inc., 478 F.3d 413, 421 (1st Cir. 2007) (“We note that liability under the ECPA is specifically exempted from Section 230 immunity.”). 281 See generally CRS Report R41733, Privacy: An Overview of the Electronic Communications Privacy Act, by Charles Doyle. 282 See 47 U.S.C. § 230(e)(1). 283 See id. § 230(e)(4). 284 18 U.S.C. §§ 2511 (outlining the prohibitions), 2520 (authorizing civil suits). 285 Doe v. GTE Corp., 347 F.3d 655, 662 (7th Cir. 2003). 286 Id. at 658–59. 287 Allow States and Victims to Fight Online Sex Trafficking Act of 2017, Pub. L. No. 115-164, § 4, 132 Stat. 1253, 1254 (2018). 288 The statute refers to civil actions “brought under” 18 U.S.C. § 1595. 47 U.S.C. § 230(e)(5)(A). See also J.B. v. G6 Hosp., LLC, No. 19-cv-07848-HSG, 2020 WL 4901196, at *7 (N.D. Cal. Aug. 20, 2020) (concluding that this exception does not include “state law civil sex trafficking claims”). At least one federal court has concluded that Section 230 should also allow civil actions brought under 18 U.S.C. § 2421A(c), which was passed as part of FOSTA. Id. at *6. 289 47 U.S.C. § 230(e)(5). 290 See id.

Congressional Research Service 28 Section 230: An Overview

plaintiff had to prove a violation of the criminal law, 18 U.S.C. § 1591, in order to avoid Section 230 immunity.291 The plaintiff was a minor who had been “convinced” by adults to send and receive sexually explicit pictures over an online messaging service.292 She sued the messaging service, arguing that the company knew or should have known that its services had been used this way and should have implemented policies to prevent this use.293 The court noted first that the lawsuit would be barred by Section 230 if an exception did not apply, because it would treat an interactive computer service provider (the messaging service) “as though [it] published and solicited the photographs” provided by others.294 Consequently, in the court’s view, the suit could proceed only if the underlying conduct violated 18 U.S.C. § 1591.295 In that case, however, the parties disputed the precise elements of the claim; specifically, they disputed the mens rea, or mental state, the plaintiff had to show to qualify for the FOSTA exception to Section 230.296 The plaintiff argued that because she was bringing a civil suit, she had to show only that the service “‘knew or should have known’ that it was participating in a venture that was engaged in sex trafficking.”297 She claimed that the federal law authorizing civil lawsuits did not incorporate any heightened mental state beyond this constructive knowledge standard.298 The trial court disagreed, ruling instead that the plaintiff had to meet the mental state standard of the criminal statute, which required “knowing and active participation in sex trafficking.”299 The court pointed to the language of the FOSTA exception, which applies to a claim only “if the conduct underlying the claim constitutes a violation of [18 U.S.C. §] 1591,” the criminal statute.300 The court then concluded that the plaintiff had not met this bar, because she had “not alleged facts that would plausibly establish that [the service] knowingly participated in the sex trafficking venture involving her,” and dismissed the claim.301

Reform Proposals and Considerations for Congress This section of the report discusses various proposals to reform Section 230, as well as some of the legal considerations implicated by those proposals, including looking to the FCC’s authority to regulate in this area and relevant First Amendment issues.

291 Doe v. Kik Interactive, Inc., No. 20-60702, 2020 WL 5156641, at *7 (S.D. Fla. Aug. 31, 2020). See also M. L. v. Craigslist Inc., No. C19-6153, 2020 WL 5494903, at *4 (W.D. Wash. Sept. 11, 2020) (concluding that FOSTA did “not create an exemption for all § 1595 claims,” but only for violations “of § 1591”). 292 Kik Interactive, 2020 WL 5156641, at *1. 293 Id. 294 Id. at *5. 295 Id. 296 Id. at *6. 297 Id. (quoting 18 U.S.C. § 1595) (emphasis added). 298 Id. 18 U.S.C. § 1595 provides: “An individual who is a victim of a violation of this chapter may bring a civil action against the perpetrator (or whoever knowingly benefits, financially or by receiving anything of value from participation in a venture which that person knew or should have known has engaged in an act in violation of this chapter) . . . .” The trial court noted that other courts have concluded that 18 U.S.C. § 1595 requires only a constructive knowledge standard. Kik Interactive, 2020 WL 5156641, at *6. 299 Kik Interactive, 2020 WL 5156641, at *7. Specifically, 18 U.S.C. § 1591(e)(5) defines “participation in a venture” to mean “knowingly assisting, supporting, or facilitating a violation.” 300 Kik Interactive, 2020 WL 5156641, at *7 (quoting 47 U.S.C. § 230(e)(5)(A)) (emphasis omitted). 301 Id.

Congressional Research Service 29 Section 230: An Overview

Overview of Reform Proposals Following the enactment of FOSTA in 2018, the push to reform Section 230 gained further momentum in Congress. Twenty-six bills in the 116th Congress would have amended Section 230.302 There were also a number of reform proposals from outside commentators and the executive branch. On May 28, 2020, President Trump issued the Executive Order on Preventing Online Censorship outlining the executive branch’s position on Section 230, weighing in on several interpretive disputes, and directing a few different agencies to take certain actions to implement those understandings.303 In response to this order, in July 2020, the National Telecommunications and Information Administration (NTIA) filed a petition for rulemaking with the FCC, the agency generally responsible for administering the Communications Act of 1934.304 The petition asked the FCC “to clarify ambiguities in [S]ection 230.”305 In October 2020, the Chairman of the FCC announced that he “intend[ed] to move forward with a rulemaking,”306 although the FCC did not act on the petition prior to the Chairman’s departure in January 2021.307 The Department of Justice also announced its own views on Section 230 in June 2020, sending proposed amendments to Congress in September 2020.308 While some Members of Congress have proposed to repeal Section 230 entirely,309 others suggest more incremental rollbacks, removing immunity only for certain types of claims or for certain providers. For instance, a number of bills would have created new exceptions to Section 230 carving out certain categories of claims, similar to FOSTA.310 A few bills focused on child sexual exploitation, allowing claims premised on conduct that violates new or existing laws related to

302 See, e.g., The Telecommunication Act’s “Good Samaritan” Protection: Section 230, DISRUPTIVE COMPETITION PROJECT, https://www.project-disco.org/section-230/#230proposals (last updated Mar. 24, 2021); see also CRS Report R46662, Social Media: Misinformation and Content Moderation Issues for Congress, by Jason A. Gallo and Clare Y. Cho, Appendix B. Although this report focuses largely on bills introduced in the 116th Congress, Section 230 repeal and reform bills have been introduced in the 117th Congress. See, e.g., H.R. 874, 117th Cong. (2021); SAFE TECH Act, S. 299, 117th Cong. (2021). 303 Exec. Order No. 13,925, 85 Fed. Reg. 34,079 (June 2, 2020). For a more in-depth discussion of this order, see CRS Legal Sidebar LSB10484, UPDATE: Section 230 and the Executive Order on Preventing Online Censorship, by Valerie C. Brannon et al. 304 NTIA, Petition for Rulemaking (July 27, 2020), https://ecfsapi.fcc.gov/file/10803289876764/ntia_petition_for_ rulemaking_7.27.20.pdf [hereinafter NTIA Petition]. 305 Id. at 5. 306 Statement of Chairman Pai on Section 230 (Oct. 15, 2020), https://docs.fcc.gov/public/attachments/DOC- 367567A1.pdf. 307 Chairman Pai Statement upon Departing the FCC (Jan. 20, 2021), https://docs.fcc.gov/public/attachments/DOC- 369408A1.pdf. 308 See Department of Justice’s Review of Section 230 of the Communications Decency Act of 1996, U.S. DEP’T OF JUSTICE, https://www.justice.gov/ag/department-justice-s-review-section-230-communications-decency-act-1996 (last visited Apr. 7, 2020); U.S. DEP’T OF JUSTICE, SECTION 230—NURTURING INNOVATION OR FOSTERING UNACCOUNTABILITY?: KEY TAKEAWAYS AND RECOMMENDATIONS 18 (2020), https://www.justice.gov/file/1286331/ download [hereinafter DOJ Recommendations]. 309 E.g., S. 5085, 116th Cong. (2020) (proposing to repeal Section 230); S. 5020, 116th Cong. (2020) (proposing to sunset Section 230 on January 2, 2023); H.R. 8896, 116th Cong. (2020) (proposing to repeal Section 230). 310 Holding Sexual Predators and Online Enablers Accountable Act, S. 5012, 116th Cong. § 5 (2020); Protecting Americans from Dangerous Algorithms Act, H.R. 8636, 116th Cong. § 2 (2020); PACT Act, S. 4066, 116th Cong. § 7 (2020); Stopping Big Tech’s Censorship Act, S. 4062, 116th Cong. § 2 (2020); Limiting Section 230 Immunity to Good Samaritans Act, S. 3983, 116th Cong. § 2 (2020); COOL Online Act, S. 3707, 116th Cong. § 3 (2020); EARN IT Act of 2020, S. 3398, 116th Cong. § 5 (2020); Protecting Local Authority and Neighborhoods Act, H.R. 4232, 116th Cong. § 2 (2019); DOJ Recommendations, supra note 308, at 18 (2020) (supporting exceptions for “facilitating terrorism” or “cyberstalking”).

Congressional Research Service 30 Section 230: An Overview

distributing child sexual abuse material.311 Others would have expanded the existing exemption for federal criminal prosecutions to include the enforcement of federal civil laws, whether all federal laws312 or only specific federal civil actions.313 Still other proposals would have created exemptions for certain lawsuits brought under state law, including breach of contract claims314 or claims relating to property rentals.315 At least some service providers may respond to the removal of Section 230 immunity for specific types of content by removing or restricting access to the content rather than facing the threat of a lawsuit.316 For example, Craigslist took down its personal ads section in response to FOSTA.317 Accordingly, creating new exceptions could cause service providers and users to remove or restrict content that could possibly be subject to those exceptions more frequently, either preemptively or in response to litigation. However, removing Section 230 immunity for certain types of content does not necessarily mean that a provider or user will be liable for hosting that content; it merely means that such liability will not be barred by Section 230. It is possible providers could continue to host that content if they believe the benefits of hosting such content outweigh potential litigation costs, particularly if providers believe they are likely to prevail in any lawsuits or that lawsuits are unlikely. Other proposals would have more broadly exposed providers to liability for hosting unlawful content, if the provider is aware of that content.318 For example, the Platform Accountability and

311 E.g., Holding Sexual Predators and Online Enablers Accountable Act, S. 5012, 116th Cong. § 5 (2020) (creating a new exception for civil actions and state criminal prosecutions that would violate new criminal offense, 18 U.S.C. § 2260B, relating to child sexual exploitation materials); EARN IT Act of 2020, S. 3398, 116th Cong. § 5 (2020) (creating new exceptions for certain civil actions and state criminal prosecutions related to specified child sexual exploitation offenses). 312 E.g., PACT Act, S. 4066, 116th Cong. § 7 (2020) (providing that Section 230 does not apply to the enforcement of federal civil statutes or regulations); Stopping Big Tech’s Censorship Act, S. 4062, 116th Cong. § 2 (2020) (creating a new exception for civil enforcement actions brought by the federal government arising out of violations of federal law). 313 E.g., Protecting Americans from Dangerous Algorithms Act, H.R. 8636, 116th Cong. (2020) (providing that “an interactive computer service shall be considered to be an information content provider” and will not receive immunity under Section 230(c)(1) in civil actions brought under 42 U.S.C. §§ 1985, 1986, or 18 U.S.C. § 2333, if the claim involves the use of an algorithm to deliver the relevant content, with certain exceptions); COOL Online Act, S. 3707, 116th Cong. § 3 (2020) (making it unlawful, “[n]otwithstanding . . . section 230(c)(1) of the Communications Act of 1934,” to make “any false or deceptive representation” that a product is of United States origin). 314 Limiting Section 230 Immunity to Good Samaritans Act, S. 3983, 116th Cong. § 2 (2020) (creating a new exception for claims “for breach of contract, promissory estoppel, or breach of a duty of good faith”). 315 Protecting Local Authority and Neighborhoods Act, H.R. 4232, 116th Cong. § 2 (2019) (creating a new exception for state law civil claims if “(A) the claimant alleges such provider facilitated the lease or rental of real property in a circumstance in which a law or contractual agreement restricts such lease or rental; (B) the claimant provides written notice of the alleged violation to such provider; and (C) such provider fails to cure the alleged violation within 30 days after the date on which such provider receives such notice”). 316 See, e.g., Corynne McSherry, Ph.D., Legal Director, Electronic Frontier Foundation, Statement for the Department of Justice Public Workshop: Section 230—Nurturing Innovation or Fostering Unaccountability? 4 (Feb. 19, 2020), https://www.justice.gov/file/1286206/download#page=82 (stating that in response to FOSTA, online companies “increased restrictions on speech discussing sex,” citing Craigslist and Tumblr as two examples). 317 See Brian Feldman, Craigslist’s Legendary Personals Section Shuts Down, N.Y. MAG. (Mar. 23, 2018), https://nymag.com/intelligencer/2018/03/craigslist-shuts-down-personals-section-because-of-fosta.html. Craigslist expressly cited FOSTA as the motive for its decision; others have speculated that Reddit and Tumblr, among other sites, made changes to their content policies in response to FOSTA. See, e.g., Paris Martineau, Tumblr’s Porn Ban Reveals Who Controls What We See Online, WIRED (Dec. 4, 2018), https://www.wired.com/story/tumblrs-porn-ban- reveals-controls-we-see-online. 318 E.g., See Something, Say Something Online Act of 2020, S. 4758, 116th Cong. § 5 (2020) (providing that a provider “that fails to report a known suspicious transmission may be held liable as a publisher for the . . . transmission”); PACT

Congressional Research Service 31 Section 230: An Overview

Consumer Transparency Act (PACT Act) would have amended Section 230 so that certain providers lose immunity under subsection (c)(1) if the provider is notified about illegal content or activity occurring on its service and “does not remove the illegal content or stop the illegal activity within 24 hours.”319 The PACT Act also would have created a procedure for notifying providers about illegal content.320 Like some other proposals, the PACT Act would have differentiated between different types of providers: small business providers would be exempt from the 24-hour deadline, and the new provisions would not apply at all to “internet infrastructure services” such as web-hosting services.321 The PACT Act would have stood in contrast to the notice-and-takedown procedures of the Digital Millennium Copyright Act (DMCA).322 The DMCA provides a “safe harbor” to covered providers who remove content after being notified that the content may violate federal copyright law, protecting them from lawsuits premised on hosting potentially infringing content.323 The European Union’s e-Commerce Directive operates somewhat similarly, providing that for certain service providers to receive immunity, the provider must “act expeditiously to remove or to disable access to the information concerned” once the service “obtain[s] actual knowledge or awareness of illegal activities.”324 However, the scope of the intermediary liability provisions in both the DMCA and the e-Commerce Directive have also been subject to debate and proposals for reform.325 One issue in proposals that would have conditioned immunity on removing unlawful content is who determines whether the content is unlawful, and how. Both the DMCA and the e-Commerce Directive essentially leave the initial determination of whether content is illegal to private parties. Under the DMCA, the person notifying a service provider of copyright infringement must submit a statement under penalty of perjury identifying the allegedly infringing material and providing a good-faith assertion that the use of the material is unlawful.326 The notifier thus has the initial burden of discovering the material and determining whether it violates copyright laws.327 The provider hosting the allegedly infringing content then must decide whether to accept the notice

Act, S. 4066, 116th Cong. § 6 (2020) (providing that Section 230(c)(1) will not apply to illegal content or activity if covered providers know about the illegal content and do not remove it within 24 hours). 319 PACT Act, S. 4066, 116th Cong. § 6 (2020). 320 Id. The bill defines “illegal activity” as content provider activity “that has been determined by a Federal or State court to violate Federal criminal or civil law.” Id. It defines “illegal content” as information that a Federal or State court has determined violates “(A) Federal criminal or civil law; or (B) State defamation law.” Id. 321 Id. A provider counts as a small business provider if, during the prior two years, it “(A) received fewer than 1,000,000 monthly active users or monthly visitors; and (B) accrued revenue of less than $25,000,000.” Id. § 2(9). 322 17 U.S.C. § 512(c); see also, e.g., CRS Report R43436, Safe Harbor for Online Service Providers Under Section 512(c) of the Digital Millennium Copyright Act, by Brian T. Yeh. It could also be compared to the notice-based liability imposed on distributors of defamatory content. See supra notes 59 to 61 and accompanying text. 323 See generally CRS In Focus IF11478, Digital Millennium Copyright Act (DMCA) Safe Harbor Provisions for Online Service Providers: A Legal Overview, by Kevin J. Hickey; U.S. COPYRIGHT OFFICE, SECTION 512 OF TITLE 17 (2020), https://www.copyright.gov/policy/section512/section-512-full-report.pdf. 324 Parliament and Council Directive 2000/31/EC, 2000 O.J. (L 178) 6. 325 See, e.g., Britain Eakin, Draft Bill That Would Revamp DMCA Sparks Divided Reaction, Law360 (Dec. 22, 2020), https://www.law360.com/publicpolicy/articles/1340196/draft-bill-that-would-revamp-dmca-sparks-divided- reaction?nl_pk=77a8fbcd-0ce9-4d0f-a0ac-3a4c7fd100a8; Illegal Content on Online Platforms, EUROPEAN COMM’N, https://ec.europa.eu/digital-single-market/en/digital-services-act-package (last updated Mar. 3, 2021). 326 17 U.S.C. § 512(c)(3). 327 See, e.g., Lenz v. Univ. Music Corp., 815 F.3d 1145, 1151 (9th Cir 2016) (noting the copyright holder’s obligation to state that the use is unauthorized and holding that this provision requires the holder to consider whether the potentially infringing material is authorized as “fair use” of a copyright).

Congressional Research Service 32 Section 230: An Overview

and expeditiously take down the material, or instead to ignore the notice and risk liability. In some ways, the DMCA incentivizes removals by granting immunity to providers that remove the allegedly infringing material, creating the risk that providers will take down lawful material rather than risk litigation.328 However, the DMCA also provides a process for the user who posted the allegedly infringing material to challenge the initial notice.329 If there is such a “counter notification,” the provider may be able to replace the initial post and retain immunity.330 Congress took a different approach in Section 230,331 insulating providers from liability for hosting both lawful and unlawful third-party content even if the provider has notice of allegedly unlawful user-generated content.332 If Congress were to decide that this approach is no longer appropriate, it might consider whether to leave the analysis of lawfulness largely to private parties, retaining some immunity for providers, as in the DMCA, or whether instead to leave the analysis to courts. In comparison, some proposals, including the PACT Act, would have imposed notice liability on providers only if a court has already adjudicated the content to be unlawful.333 Proposals like the PACT Act would state that if a provider is aware of an unlawful post, it will lose immunity for a lawsuit premised on that specific post.334 By contrast, a number of other proposals would have stipulated that service providers or users cannot claim Section 230 immunity if they engage in certain conduct, seemingly regardless of whether a subsequent lawsuit is premised on that conduct. For example, the Stopping Big Tech’s Censorship Act would have provided that both service providers and users may only claim immunity under Section 230(c)(1) if a service “takes reasonable steps to prevent or address the unlawful use” of the service “or the unlawful publication of information on” the service.335 It appears that this proposal would have prevented both providers and users from claiming Section 230 immunity in a lawsuit regardless of whether the service acted reasonably with respect to content that is the basis of a specific lawsuit.336 The CASE-IT Act would have taken a similar approach, providing that service providers and users lose Section 230(c)(1) immunity for a year if they engage in certain activities, including permitting harmful content to be distributed to minors, if the harmful content “is made readily accessible to minors by the failure of such provider or user to implement a system

328 17 U.S.C. § 512(g)(1); see also, e.g., Wendy Seltzer, Free Speech Unmoored in Copyright’s Safe Harbor: Chilling Effects of the DMCA on the First Amendment, 24 HARV. J. LAW & TECH. 171, 175 (2010) (discussing the incentive structure and arguing that the DMCA results in removal of constitutionally protected speech). 329 17 U.S.C. § 512(g)(2)–(3). 330 Id. § 512(g)(2), (4). 331 Both provisions, however, adopt a policy of immunity for intermediaries hosting others’ content, and grant significant power to private parties to determine what content should be online. See, e.g., Markham C. Erickson & Sarah K. Leggin, Exporting Internet Law Through International Trade Agreements: Recalibrating U.S. Trade Policy in the Digital Age, 24 CATH. U. J. L. & TECH. 317, 340 (2016). 332 Cf., e.g., Barrett v. Rosenthal, 146 P.3d 510, 520 (Cal. 2006) (comparing the DMCA’s “limited liability” scheme to Section 230, and concluding “that Congress did not intend to permit notice liability under the CDA”). However, as noted above, Section 230 immunity contains exceptions allowing liability for hosting certain types of unlawful content, including if a site violates federal criminal law. Supra “Section 230(e): Exceptions.” 333 PACT Act, S. 4066, 116th Cong. § 6 (2020). See also DOJ Recommendations, supra note 308, at 18 (“Section 230 should be narrowed so as not to apply in actions where a platform has failed to take down content or activity, within a reasonable time, after receiving notice that a court in the United States has adjudicated the content or activity to be unlawful.”). 334 See PACT Act, S. 4066, 116th Cong. § 6 (2020). 335 Stopping Big Tech’s Censorship Act, S. 4062, 116th Cong. § 2 (2020). 336 See id.

Congressional Research Service 33 Section 230: An Overview

designed to effectively screen users who are minors from accessing such content.”337 In the same vein, other bills would have caused providers to lose Section 230 immunity if they use algorithms to distribute content to users or display behavioral advertising.338 Some efforts focus more fundamentally on what activities qualify someone as an information content provider, seeking to clarify or roll back court decisions that sponsors believe misinterpret the term “information content provider” and misapply Section 230(c)(1).339 Two proposals introduced in the 116th Congress would have treated a person as an information content provider if the person “affirmatively and substantively” modifies another’s content.340 One of the bills would have also included people who solicit or fund information.341 In contrast to proposals that would expose service providers or users to greater liability for hosting or distributing another’s content, a number of other proposals would have limited providers’ and users’ immunity for restricting access to content. Some have suggested clarifying that Section 230(c)(1) applies only when a suit is premised on hosting content, and Section 230(c)(2) is the sole provision that immunizes takedown decisions.342 This would resolve the interpretive dispute mentioned above regarding the scope of Section 230(c)(1) and its relationship to (c)(2).343 Significantly, because of this interpretive dispute, if a proposal amends only Section 230(c)(2) and does not address the scope of Section 230(c)(1), it is possible that courts would continue to apply (c)(1) to takedown decisions regardless of whether the more limited immunity in (c)(2) protects those decisions. Accordingly, to limit immunity for decisions to restrict access to content, the scope of Section 230(c)(1) may have to be narrowed.

337 CASE-IT Act, H.R. 8719, 116th Cong. § 2 (2020). 338 Break Up Big Tech Act of 2020, H.R. 8922, 116th Cong. § 2 (2020) (providing that Sections 230(c)(1) and (2) will not apply if a provider (1) sells targeted advertising and displays the advertising to users who have not opted in; (2) “places items, or facilitates the placement of items, into the stream of commerce”; (3) collects data for commercial purposes; or (4) “uses a design or product that addicts . . . users”); Don’t Push My Buttons Act, S. 4756, 116th Cong. § 2 (2020) (providing that a provider generally may not claim immunity under Section 230(c)(1) or (2) if the provider uses automated functions to deliver content to users based on information it has collected about the user’s habits, preferences, or beliefs, with certain exceptions); BAD ADS Act, S. 4337, 116th Cong. § 2 (2020) (preventing certain providers from claiming immunity under Section 230 for 30 days after displaying “behavioral advertising” to a user or providing user information to another person knowing that such information will be used to “create or display behavioral advertising”). Cf. Biased Algorithm Deterrence Act of 2019, H.R. 492, 116th Cong. § 2 (2019) (stating that if “an owner or operator of a social media service . . . displays user-generated content in an order other than chronological order, delays the display of such content relative to other content, or otherwise hinders the display of such content relative to other content, if for a reason other than to restrict access to or availability of material described in [Section 230(c)(2)(A)] or to carry out the direction of the user that generated such content,” that social media service “shall be treated as a publisher or speaker of such content”). 339 Cf., e.g., NTIA Petition, supra note 304, at 42–46 (discussing cases and outlining circumstances under which a service provider should be deemed to have published content). 340 Protect Speech Act, H.R. 8517, 116th Cong. § 2 (2020) (providing that a person qualifies as an “information content provider” if the person “solicits, comments upon, funds, or affirmatively and substantively contributes to, modifies, or alters information”); Online Freedom and Viewpoint Diversity Act, S. 4534, 116th Cong. § 2 (2020) (amending the definition of “information content provider” to include “any instance in which a person or entity editorializes or affirmatively and substantively modifies the content of another person or entity”). See also NTIA Petition, supra note 304, at 42 (clarifying that the statutory definition of “information content provider” includes “substantively contributing to, modifying, altering, presenting or prioritizing with a reasonably discernible viewpoint, commenting upon, or editorializing about content provided by another information content provider”). 341 Protect Speech Act, H.R. 8517, 116th Cong. § 2 (2020). 342 Protect Speech Act, H.R. 8517, 116th Cong. § 2 (2020); Online Freedom and Viewpoint Diversity Act, S. 4534, 116th Cong. § 2 (2020); NTIA Petition, supra note 304, at 31. 343 Supra notes 75 to 76 and accompanying text.

Congressional Research Service 34 Section 230: An Overview

Some proposals would have amended the categories of material mentioned in Section 230(c)(2), changing the types of content that providers and users may restrict with immunity.344 For example, a few proposals would delete the catch-all term granting immunity for “otherwise objectionable material” and replace this with new, more limited categories of material such as material “promoting terrorism” or “promoting self-harm,” or “unlawful” material.345 In terms of subject matter, the category of “unlawful” material is broader than material that promotes terrorism or fits within one of the other specific categories of material mentioned in Section 230(c)(2)—but the category is also narrower in the sense that, for example, not all material “promoting terrorism” may be unlawful.346 Using the phrase “unlawful” makes these proposals subject to the same questions discussed above regarding who determines whether the content is unlawful and how.347 Other new categories would likely also raise interpretive questions, such as whether specific material promotes terrorism or self-harm. If a proposal retains the language in Section 230(c)(2) providing immunity for restricting material “that the provider or user considers to” fall within the listed categories,348 it is likely that courts would continue interpreting this provision as embodying a largely subjective standard.349 Some proposals, though, would have amended Section 230(c)(2) to state that this provision applies only if the provider or user has an “objectively reasonable” belief that the content falls within one of the listed categories,350 seemingly inviting courts to engage in a more rigorous review of this belief.351 A number of other proposals would have limited immunity for takedown decisions in ways that depart more substantially from the current Section 230(c)(2) framework. For example, a few proposals would have granted immunity for takedown decisions only if the provider or user has acted in a viewpoint-neutral manner, meaning in a way that is not biased against any particular viewpoint.352 Other proposals would require service providers or users to adopt certain procedural

344 See, e.g., Stop the Censorship Act, H.R. 4027, 116th Cong. § 2 (2019) (replacing entire list of adjectives in Section 230(c)(2) with “unlawful”). 345 Stop Suppressing Speech Act of 2020, S. 4828, 116th Cong. § 2 (2020) (replacing “harassing, or otherwise objectionable” with “or harassing, that the provider or user determines to be unlawful, or that promotes violence or terrorism”); Protect Speech Act, H.R. 8517, 116th Cong. § 2 (2020) (replacing “harassing, or otherwise objectionable” with “promoting terrorism or violent extremism, harassing, promoting self-harm, or unlawful”); Online Freedom and Viewpoint Diversity Act, S. 4534, 116th Cong. § 2 (2020) (replacing “otherwise objectionable” with “promoting self- harm, promoting terrorism, or unlawful”); Stop the Censorship Act of 2020, H.R. 7808, 116th Cong. § 2 (2020) (replacing “otherwise objectionable” with “unlawful, or that promotes violence or terrorism”). 346 Cf. Holder v. Humanitarian Law Project, 561 U.S. 1, 25–26, 39 (2010) (ruling that some speech related to terrorism may be protected by the First Amendment, including political advocacy, but upholding the constitutionality of a law criminalizing “material support” to terrorists under a strict scrutiny analysis). 347 See supra notes 318 to 333 and accompanying text. 348 47 U.S.C. § 230(c)(2)(A). 349 See supra notes 209 to 212 and accompanying text. 350 See Protect Speech Act, H.R. 8517, 116th Cong. § 2 (2020); Online Freedom and Viewpoint Diversity Act, S. 4534, 116th Cong. § 2 (2020). 351 Cf. Holomaxx Techs. v. Microsoft Corp., 783 F. Supp. 2d 1097, 1104 (N.D. Cal. 2011) (“No court has articulated specific, objective criteria to be used in assessing . . . a provider’s subjective determination of what is ‘objectionable’ . . . . Here, however, it is clear . . . that Microsoft reasonably could conclude that Holomaxx’s emails were ‘harassing’ and thus ‘otherwise objectionable.’” (emphasis added)). 352 See, e.g., CASE-IT Act, H.R. 8719, 116th Cong. § 2 (2020) (providing that Sections 230(c)(1) and (c)(2)(A) will not apply to certain providers if the provider “makes content moderation decisions pursuant to policies or practices that are not reasonably consistent with the First Amendment to the Constitution,” treating these providers as equivalent to state actors); Stopping Big Tech’s Censorship Act, S. 4062, 116th Cong. § 2 (2020) (providing that Section 230(c)(2)(A) will apply only if a provider or user, among other requirements, acts “in a viewpoint-neutral manner”); Ending Support

Congressional Research Service 35 Section 230: An Overview

safeguards or notice requirements related to takedowns in order to benefit from Section 230 immunity: for instance, by stating that providers must adhere to their terms of service in restricting access to user content.353 Some of these proposals would have tied these requirements to the “good faith” standard in Section 230(c)(2)(A), stating that before a provider may be considered to be acting in good faith, it must clearly state its criteria for restricting access to content and then adhere to that criteria in individual decisions.354 At least two bills would have also required users to provide explanations of their decisions to restrict access to content.355 Although there have been many proposals to reform Section 230’s immunity shield, some argue either that Section 230 should not be changed or that reforms should be modest and carefully considered.356 And as commentators have noted, some of the reform proposals may conflict with others and pursue divergent goals, making it more difficult to predict which of these reform efforts, if any, may garner sufficient support.357

FCC Congress passed the CDA as part of the Telecommunications Act of 1996. The Telecommunications Act, in turn, amended the Communications Act of 1934, a statute administered by the FCC. Accordingly, the FCC may have a role in enforcing or implementing Section 230, although Section 230 does not explicitly mention the FCC, and the FCC has not interpreted Section 230 since its passage.358 The FCC has, on occasion, referred to Section 230 in for Internet Censorship Act, S. 1914, 116th Cong. § 2 (2019) (providing that Section 230(c) will not apply to larger providers unless the FTC has certified that “the company does not moderate information . . . in a manner that is biased against a political party, political candidate, or political viewpoint”). 353 E.g., Protect Speech Act, H.R. 8517, 116th Cong. § 2 (2020) (providing that Section 230(c)(1) and Section 230(c)(2)(A) apply only if the provider or user (1) makes publicly available terms of service for content moderation; (2) restricts content consistently with those terms of service; (3) does not restrict content “on deceptive grounds or apply terms of service or use to restrict access to or availability of material that is similarly situated to material that the service intentionally declines to restrict”; and (4) gives the content provider “timely notice” of the basis for restricting access to the content and “a meaningful opportunity to respond”); Stopping Big Tech’s Censorship Act, S. 4062, 116th Cong. § 2 (2020) (providing that to claim Section 230(c) immunity, (1) a provider must “clearly explain” its “practices and procedures” for restricting access to material; and (2) a provider or user who restricts access to material must “provide a clear explanation of that decision” to the content provider); Limiting Section 230 Immunity to Good Samaritans Act, S. 3983, 116th Cong. § 2 (2020) (amending Section 230(c)(1) so that it applies to a covered “edge provider” only if it adopts written terms of service for restricting material that “promise” that the provider will (1) “design and operate” the service in “good faith,” a defined term, and (2) pay certain damages and costs if the provider is found to have breached that promise). 354 See, e.g., NTIA Petition, supra note 304, at 39; DOJ Recommendations, supra note 308, at 22. 355 Protect Speech Act, H.R. 8517, 116th Cong. § 2 (2020); Stopping Big Tech’s Censorship Act, S. 4062, 116th Cong. § 2 (2020). 356 See, e.g., Clyde Wayne Crews, The Case Against Social Media Content Regulation, COMPETITIVE ENTER. INST. (June 1, 2020), https://cei.org/issue_analysis/the-case-against-social-media-content-regulation; Section 230 of the Communications Decency Act, ELEC. FRONTIER FOUND., https://www.eff.org/issues/cda230 (last visited Apr. 7, 2021) Eric Goldman, Want to Learn More About Section 230? A Guide to My Work, TECH. & MKTG. L. BLOG (July 1, 2020), https://blog.ericgoldman.org/archives/2020/07/want-to-learn-more-about-section-230-a-guide-to-my-work.htm; Jennifer Huddleston, Does Content Moderation Need Changes to Section 230?, AM. ACTION FORUM (June 18, 2020), https://www.americanactionforum.org/insight/does-content-moderation-need-changes-to-section-230/. 357 See, e.g., Dean DeChiaro, OK to Change Section 230, Tech CEOs Say, But How Remains Elusive, CONG. Q (Nov. 17, 2020), https://plus.cq.com/doc/news-6052674; Jonathan Grieg, Section 230 Hangs in the Balance After Attacks from Biden and Trump, TECHREPUBLIC (Nov. 4, 2020), https://www.techrepublic.com/article/section-230-hangs-in-the- balance-after-attacks-from-biden-and-trump. 358 In past regulatory actions involving broadband internet access service, the FCC has explicitly declined to regulate internet services other than those relating to transmission. E.g., Protecting and Promoting the Open Internet, 30 FCC

Congressional Research Service 36 Section 230: An Overview

rulemaking proceedings, though only in a narrow context: specifically, the FCC has repeatedly cited the policy statement in Section 230(b)—not Section 230(c)—of “preserv[ing] the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation.”359 The D.C. Circuit has held, however, that the FCC may not rely on Section 230(b) as a source of regulatory authority.360 Because the FCC has not issued regulations interpreting the scope of Section 230 immunity, the FCC’s rulemaking authority over Section 230 remains unclear. As discussed above,361 the NTIA Petition (and comments submitted in response) provided the FCC with an opportunity to consider its rulemaking authority, and some FCC communications support the notion that the agency has authority to issue regulations interpreting Section 230.362 Although the FCC may decline to act on the NTIA Petition, which would leave the issue of the FCC’s legal authority over Section 230 unresolved, this section will nonetheless explore the possible legal issues an FCC decision to regulate may raise. If the FCC does not have jurisdiction to regulate, Congress could vest the FCC with regulatory authority over Section 230 through legislation.

Existing Legal Authority The NTIA Petition is the first rulemaking petition involving Section 230(c) to appear on the FCC’s docket. Because the FCC may not regulate absent congressional authorization to do so, 363 the NTIA Petition and comments responding to the petition extensively explore the agency’s authority to issue rules. The FCC exercises regulatory authority in two instances. First, the agency may exercise regulatory authority to advance specific statutory objectives: for example, Section 201 of the Communications Act prohibits certain telecommunications service providers from engaging in “unjust or unreasonable” practices,364 and the FCC has relied on this provision to introduce numerous regulations classifying particular practices as unjust or unreasonable.365 Second, the

Rcd. 5601, 5775, para. 382 (2015) (noting that the FCC does not intend to regulate “any internet applications or content.”). 359 E.g., Implementation of Section 621(a) of the Cable Communications Policy Act, 34 FCC Rcd. 6844, 6899, para. 102 (2019); Restoring Internet Freedom, 33 FCC Rcd. 311, 432, para. 203 (2018). 360 Mozilla Corp. v. FCC, 940 F.3d 1, 78 (D.C. Cir. 2019). The Mozilla court did not address whether the FCC could otherwise issue regulations interpreting Section 230 while relying on a different source of authority. See “Existing Legal Authority,” infra, for more discussion of this issue. 361 See supra “Overview of Reform Proposals.” 362 Statement of Chairman Pai on Section 230 (Oct. 15, 2020), https://docs.fcc.gov/public/attachments/DOC- 367567A1.pdf; see Thomas M. Johnson Jr., The FCC’s Authority to Interpret Section 230 of the Communications Act, FCC.GOV (Oct. 21, 2020), https://www.fcc.gov/news-events/blog/2020/10/21/fccs-authority-interpret-section-230- communications-act. But cf. Thomas M. Johnson Jr. and Michael Carson, Caveat Lector: A Blog Post About Reliance on Agency Blog Posts, FCC.GOV (Apr. 4, 2019), https://www.fcc.gov/news-events/blog/2019/04/04/caveat-lector-blog- post-about-reliance-agency-blog-posts (explaining that blog posts are not authoritative statements of law or the agency’s position). In spite of former Chairman Pai’s statement, the FCC has not acted on the NTIA Petition, and it appears unlikely to do so. See Emily Birnbaum, Ajit Pai Is Distancing Himself from President Trump, PROTOCOL (Jan. 7, 2021), https://www.protocol.com/ajit-pai-distancing-trump. 363 See La. Pub. Serv. Comm’n v. FCC, 476 U.S. 355, 374 (1986) (“[A]n agency literally has no power to act . . . unless and until Congress confers power upon it.”). 364 47 U.S.C. § 201. 365 E.g., 47 C.F.R. §§ 64.6000 et seq. (regulating practices relating to telecommunications services used by incarcerated people); id. §§ 64.2400 et seq. (setting standards for telecommunications service billing).

Congressional Research Service 37 Section 230: An Overview

FCC may regulate absent a specific statutory objective when its regulations are “reasonably ancillary to [its] responsibilities.”366 For the FCC to exercise this ancillary authority, it must be acting pursuant to the general grant of jurisdiction included in Title I of the Communications Act, which gives the FCC jurisdiction over “all interstate and foreign communication by wire or radio.”367 The FCC generally exercises ancillary authority to regulate services and entities not explicitly subject to regulation in the Communications Act.368 Examples of such FCC regulation include the early regulation of cable television369 and the regulation of Voice over Internet Protocol (VoIP) providers.370 With respect to the FCC’s authority to promulgate rules interpreting Section 230, the NTIA Petition and its comments largely ignore ancillary authority and instead focus on whether the agency may regulate based on its express statutory authority. The NTIA Petition argues that this authority comes not from Section 230, but from Section 201(b) of the Communications Act, which gives the FCC power to “prescribe such rules and regulations as may be necessary in the public interest to carry out” the Communications Act.371 Because Section 230 is part of the Communications Act, the NTIA petition argues, the power granted in Section 201(b) should permit the FCC to issue regulations interpreting Section 230.372 To support this position, the NTIA Petition relies on the Supreme Court’s decision in AT&T Corp. v. Iowa Utilities Board, a case that involved the FCC’s authority to regulate under certain provisions added to the Communications Act by the Telecommunications Act of 1996.373 In relevant part, the Supreme Court held that “the grant in § 201(b) means what it says: The FCC has rulemaking authority to carry out the ‘provisions of this Act,’ which include [sections] added by the Telecommunications Act of 1996.”374 The provisions at issue in Iowa Utilities Board were Sections 251 and 252 of the Communications Act, which relate to competition for local telephone service.375 Like Section 230, Sections 251 and 252 appear under Title II of the Communications Act, which generally sets forth requirements for telecommunications service providers (or “common carriers”).376 After the

366 United States v. S.W. Cable Co., 392 U.S. 157, 178 (1968); see 47 U.S.C. § 154(i) (granting the FCC power to “perform any and all acts . . . as may be necessary in the execution of its functions.”). 367 S.W. Cable Co., 392 U.S. at 178 (citing 47 U.S.C. § 152(a)). 368 Services expressly regulated by the Communications Act include telecommunications services (under Title II), radio transmission services (under Title III), and cable television services (under Title VI). See generally Comcast Corp. v. FCC, 600 F.3d 642, 645 (D.C. Cir. 2010) (setting forth the contours of the FCC’s express authority). 369 See S.W. Cable Co., 392 U.S. at 178. Although the Communications Act as originally enacted did not specifically provide for regulation of cable television, which did not exist at the time of the Act’s passage, subsequent amendments addressed cable television. E.g., Cable Communications Policy Act, Pub. L. No. 98-549, 98 Stat. 2779 (1984); Cable Television Consumer Protection and Competition Act, Pub. L. No. 102-385, 106 Stat. 1460 (1992). 370 E.g., IP-Enabled Services, E911 Requirements for IP-Enabled Service Providers, 20 FCC Rcd. 10245 (2005) (requiring VoIP providers to supply 911 emergency calling capabilities); Universal Service Contribution Methodology, 21 FCC Rcd. 7518 (establishing universal service contribution obligations for VoIP providers); Implementation of the Telecommunications Act of 1996, 22 FCC Rcd. 6927 (2007) (extending consumer privacy requirements to VoIP providers); IP-Enabled Services, 22 FCC Rcd. 11275 (2007) (extending Telecommunications Relay Service requirements to VoIP providers). 371 47 U.S.C. § 201(b). 372 NTIA Petition, supra note 304, at 16. 373 AT&T Corp. v. Iowa Utils. Bd., 525 U.S. 366 (1999). 374 Id. at 378. 375 47 U.S.C. §§ 251, 252. 376 “Common carriers” under the Communications Act are generally entities that provide “telecommunications service,” which the Act defines as “the offering of telecommunications for a fee directly to the public, or to such classes

Congressional Research Service 38 Section 230: An Overview

FCC issued rules under Sections 251 and 252, several telephone carriers and state utility commissions challenged the FCC’s authority to issue rules, which the challengers averred belonged solely to the states.377 The Supreme Court held that Section 201(b) gives the FCC broad power to make rules implementing provisions of the Communications Act, independent of whether the provisions in question grant any rulemaking authority.378 Commenters disputing that Section 201(b) gives the FCC authority to regulate under Section 230 rely on several differences between Sections 251 and 252 on the one hand and Section 230 on the other. First, the provisions of Title II of the Communications Act—including Sections 251 and 252—generally apply to common carrier providers of telecommunications service. Section 230, by contrast, applies to users and providers of interactive computer services.379 While these two categories are not, by their definitions, mutually exclusive, the definition of “interactive computer service” is broader than the definition of common carrier.380 Some commenters have argued that Section 201(b) should be interpreted to give the FCC broad rulemaking power only over common carriers otherwise subject to Title II’s provisions.381 All of Section 201(b) but its last sentence refers explicitly to common carrier service, and the cases on which NTIA relies for support interpret Section 201(b)’s rulemaking authority as applied to common carriers.382 While the holding in Iowa Utilities Board did not articulate such a limit, a dissent by Justice Breyer contended (and the majority opinion appeared to agree) that the scope of Section 201(b)’s power “to make rules implementing the more specific terms of a later enacted statute depends upon what the later statute contemplates.”383 Should a court determine that Section 230, by failing to address common carrier regulation, does not “contemplate” rulemaking under Section 201(b), the section’s general grant of authority may be unavailable. Thus, the FCC’s rulemaking authority under Section 201(b) may have limits. The NTIA Petition’s opponents also point to the lack of any language in Section 230 evincing an intent to vest the FCC with regulatory power. Section 230 as a whole does not refer to the FCC,

of users as to be effectively available directly to the public, regardless of the facilities used.” 47 U.S.C. § 153( 53). 377 Iowa Utils. Bd., 525 U.S. at 374. 378 Id. at 377–85. 379 Compare 47 U.S.C. § 153(11) with id. § 230(f)(2). Although the FCC has not spoken on whether an interactive computer service provider may be a common carrier, the agency does not treat broadband internet access service providers—a subset of interactive computer service providers—as common carriers. See Restoring Internet Freedom, 33 FCC Rcd. 311 (2018). 380 For example, the definition of interactive computer service explicitly refers to “information services,” which are distinct from common carriers. See Nat’l Cable & Telecomms. Ass’n v. Brand X Internet Servs., 545 U.S. 967, 976 (2005) (holding that “[i]nformation-service providers . . . are not subject to mandatory common-carrier regulation under Title II”). 381 E.g., Pub. Knowledge, Comments of Pub. Knowledge In re Nat’l Telecoms. & Info. Admin. Petition to “Clarify” Provisions of Section 230 of the Commc’ns Act of 1934, as Amended, at 4 (Sept. 2, 2020), https://ecfsapi.fcc.gov/file/ 109020607125130/PK_Comments_NTIA_FCC_230_Petition.pdf; Ctr. for Democracy & Tech., Comments of the Ctr. for Democracy & Tech. Opposing the Nat’l Telecoms. & Info. Admin.’s Petition for Rulemaking, at 6 (Aug. 31, 2020), https://ecfsapi.fcc.gov/file/10831957605823/CDT%20Opposition%20to%20NTIA%20Petition%20on%20Section%20 230.pdf; TechFreedom, Reply Comments of TechFreedom, at 5 (Sept. 17, 2020), https://ecfsapi.fcc.gov/file/ 10917114884764/NTIA%20230%20Petition%20Reply%20Comments%20-%209.17.2020.pdf. 382 47 U.S.C. § 201(b); Iowa Utils. Bd., 525 U.S. at 377–79; see also City of Arlington v. FCC, 569 U.S. 290, 293 (2013) (recognizing Section 201(b) rulemaking authority to promulgate an order regulating common carrier wireless services). 383 Iowa Utils. Bd., 525 U.S. at 420 (Breyer, J., dissenting); id. at 378 n.5 (majority opinion) (citing this conclusion from Justice Breyer’s dissent favorably, but noting that Congress would have been aware of Section 201(b)’s broad authority at the time of the Telecommunications Act’s passage).

Congressional Research Service 39 Section 230: An Overview

and subsection(b)(2) states Congress’s policy to keep interactive computer services “unfettered by Federal or State regulation.”384 Likewise, Section 230’s legislative history includes statements by the law’s authors expressing a desire to avoid “a Federal Computer Commission with an army of bureaucrats regulating the internet.”385 Sections 251 and 252, by comparison, provide an extensive regulatory scheme that explicitly directs the FCC to act.386 The NTIA Petition does not identify any cases in which a federal court upheld the FCC’s rulemaking authority as applied to a statutory provision that does not explicitly mention the FCC.387 Despite these distinguishing features, Section 230 is nonetheless part of Title II of the Communications Act. Though subsequent court of appeals decisions have sought specific authority justifying FCC rulemakings,388 the Supreme Court’s holding in Iowa Utilities Board is broad and includes only passing mentions of limitations on the FCC’s power to implement the provisions of the Communications Act.389 Ultimately, an FCC decision that it has authority to issue rules interpreting Section 230 would be constrained by general administrative law principles and could be given some degree of deference by a reviewing court.390

Deference to FCC Regulation An issue related to the FCC’s authority to regulate on Section 230 is the weight afforded to its regulations: when a federal agency interprets a statute, courts accord the agency’s interpretation varying levels of “deference” depending on, among other things, whether the statute evinces a congressional intent to provide the agency with regulatory authority and whether the statutory language is ambiguous.391 In determining the weight to accord an agency’s interpretation of a statute, reviewing courts engage in a multistep analysis first articulated by the Supreme Court in Chevron U.S.A., Inc. v. Natural Resources Defense Council, Inc.392 A court first inquires whether the Chevron analysis is appropriate. In United States v. Mead Corp., the Supreme Court held that agency interpretations of statutes “qualify for Chevron deference when it appears that Congress delegated authority to the agency to make rules carrying the force of law, and that the agency interpretation claiming deference was promulgated in the exercise of

384 47 U.S.C. § 230(b)(2). 385 141 CONG. REC. H8470 (daily ed. Aug. 4, 1995) (statement of Rep. Cox); see also id. at H8471 (statement of Rep. Wyden) (arguing that “the Federal Communications Commission . . . cranking out rules and proposed rulemaking programs” would be unable to keep pace with the advancement of technology). 386 E.g., 47 U.S.C. § 251(d)(1) (directing the FCC to “establish regulations to implement the requirements of this section”). 387 See NTIA Petition, supra note 304, at 15–17 (citing only Iowa Utilities Board and City of Arlington in support). 388 See, e.g., Motion Picture Ass’n of Am., Inc. v. FCC, 309 F.3d 796, 799 (D.C. Cir. 2002) (holding that FCC could not rely on general authority over television broadcasts to promulgate rules relating to television programming content). The FCC rulemaking at issue in Motion Picture Ass’n relied on statutory provisions that, like Section 230, were added as part of the Telecommunications Act of 1996 and did not direct the FCC to issue rules. See id. at 798. 389 E.g., Iowa Utils. Bd., 525 U.S. at 378 n.5. 390 See Chevron U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 842–43 (1984) (setting forth “deference” framework for administrative interpretations of statutes); City of Arlington v. FCC, 569 U.S. 290, 297 (2013) (holding that an agency’s determination of its authority to regulate is reviewed under the Chevron framework); see also FDA v. Brown & Williamson Tobacco Corp., 529 U.S. 120, 132–33 (2000) (analyzing agency exercise of regulatory authority in light of the “overall statutory scheme” when agency regulates in previously unregulated area); City of Arlington, 569 U.S. at 297 (holding that an agency’s determination of its own jurisdiction is subject to Chevron deference). 391 United States v. Mead Corp., 533 U.S. 218, 226–27 (2001); Chevron, 467 U.S. at 842–43. 392 For a more detailed review of the Chevron framework, see CRS Report R44954, Chevron Deference: A Primer, by Valerie C. Brannon and Jared P. Cole.

Congressional Research Service 40 Section 230: An Overview

that authority.”393 If the court determines Chevron analysis is not appropriate, the court may accord the agency’s interpretation “respect proportional to its ‘power to persuade’”394—i.e., the court will conduct its own interpretation and rely on the agency’s interpretation as it sees fit. If a court proceeds with the Chevron framework, it first looks at the statute at issue and determines whether the statute has left any ambiguities for the agency to address.395 Should the court determine the statute is ambiguous, it then asks whether the agency’s interpretation was reasonable, deferring to the agency interpretation if so.396 If the FCC issued rules interpreting Section 230 and those rules were challenged in court, the reviewing court may follow the Chevron analysis to determine the validity of the rules. As discussed above, a number of commenters on the NTIA petition dispute that the FCC has authority to promulgate rules interpreting Section 230.397 In light of the concerns raised by these commenters, a court reviewing an FCC rule may reason that the FCC does not have authority to “make rules carrying the force of law.”398 Even if a reviewing court chooses to apply the Chevron framework to an FCC rule, it may determine that Section 230 contains no ambiguities to be resolved—as several commenters have argued—or that the FCC’s interpretations of its own authority or of Section 230 are unreasonable.399 In short, any potential FCC rule faces several judicial hurdles. An additional wrinkle relating to judicial review of an FCC rule is that the Chevron framework generally comes into play when a party brings suit against an agency, such as the FCC, seeking judicial review of a specific agency action under the Administrative Procedure Act (APA) or an agency-specific statute providing for judicial review.400 However, Section 230 generally comes into play in litigation between private parties regarding content-related claims. While an FCC rule could be challenged in APA litigation, courts reviewing such a rule in subsequent private litigation could face questions of whether a private party may attack the validity of an FCC rule outside of the APA’s review process. In a recent case, the Supreme Court was asked whether a district court hearing a private dispute must follow an FCC order interpreting terms relevant to the dispute.401 However, the Court declined to answer this question, instead suggesting that the answer may depend on two other unraised issues: (1) whether the order set forth a “legislative rule” with the force of law,402 and (2) whether the parties had a “prior” and “adequate”

393 United States v. Mead Corp., 533 U.S. 218, 226–27 (2001). 394 Id. at 235 (quoting Skidmore v. Swift & Co., 323 U.S. 134, 140 (1944)). 395 Chevron, 467 U.S. at 842–43. 396 Id. at 843. 397 See supra “Existing Legal Authority.” 398 See Mead, 533 U.S. at 226–27. 399 See, e.g., Comput. & Commc’ns Indus. Ass’n, Comments of the Comput. & Commc’ns Indus. Ass’n, In re Section 230 of the Commc’ns Act of 1934, at 3 (Sept. 2, 2020), https://ecfsapi.fcc.gov/file/10902499105067/Corrected%20- %20CCIA%20Comments%20to%20FCC%20on%20RM-11862.pdf (“Courts have consistently held that Section 230 is clear and unambiguous . . . .”); Pub. Knowledge, Comments of Pub. Knowledge In re Nat’l Telecoms. & Info. Admin. Petition to “Clarify” Provisions of Section 230 of the Commc’ns Act of 1934, as Amended, at 2–4 (Sept. 2, 2020), https://ecfsapi.fcc.gov/file/109020607125130/PK_Comments_NTIA_FCC_230_Petition.pdf (averring that Section 230 does not delegate rulemaking authority to the FCC and “even if the agency were to pronounce upon its meaning, courts would owe it no deference”). 400 See 5 U.S.C. § 702; 47 U.S.C. § 402. 401 PDR Network, LLC v. Carlton & Harris Chiropractic, Inc., 139 S. Ct. 2051 (2019). 402 A “legislative rule” is a rule “issued by an agency pursuant to statutory authority” with the “force and effect of law.” Id. at 2055 (citing Chrysler Corp. v. Brown, 441 U.S. 281, 302–03 (1979)). An agency action that is not a “legislative rule” may instead be an “interpretive rule,” which “advis[es] the public of the agency’s construction of the statutes and

Congressional Research Service 41 Section 230: An Overview

opportunity to seek judicial review of the order, as required by the APA.403 In the wake of the Supreme Court’s decision, federal courts have accorded FCC orders varying levels of judicial deference.404 Because Section 230 may be raised as a defense against a broad array of claims, both state and federal, an FCC order interpreting terms in Section 230 may come before a multitude of different courts, and these courts in turn may not all approach the question of FCC authority in the same way.

Considerations for Congress Congress has several legislative options at its disposal to clarify the FCC’s role in administering Section 230. First, an express delegation of authority to regulate could provide the FCC with a statutory basis for promulgating regulations. Conversely, an express disavowal of authority could prohibit the FCC from attempting to regulate under Section 230. Both of these approaches could resolve the questions relating to the FCC’s statutory authority to regulate. A delegation of authority to make rules with “the force of law” would also clarify the deference courts should give any FCC rule, although reviewing courts could nonetheless reject the FCC’s interpretations based on a lack of statutory ambiguities or the unreasonableness of the FCC’s rule.405

Free Speech Considerations The Free Speech Clause of the First Amendment to the U.S. Constitution limits the government’s ability to regulate speech.406 There are at least two distinct types of First Amendment issues that may be raised by proposals to amend Section 230. The first issue is whether any given proposal unconstitutionally infringes the constitutionally protected speech of either providers or users. The second is whether, if Section 230 is repealed in whole or in part, the First Amendment may nonetheless prevent private parties or the government from holding providers liable for hosting content. This section of the report first explains background principles on legal protections for online speech, then provides some initial considerations for evaluating these two issues.

rules which it administers” and has no binding legal effect. Id. (citing Perez v. Mortgage Bankers Ass’n, 575 U.S. 92, 97 (2015) (alteration in original)). 403 Id. (citing 5 U.S.C. § 703). 404 See Carlton & Harris Chiropractic, Inc. v. PDR Network, LLC, 982 F.3d 258, 263 (4th Cir. 2020) (holding, on remand from the Supreme Court, that the FCC’s order was interpretive and interpretive rules are nonbinding on district courts); True Health Chiropractic Inc. v. McKesson Corp., No. 13-02219, 2020 WL 7664484, at *5–6 (N.D. Cal. Dec. 24, 2020) (noting that because the Supreme Court’s decision had not definitively resolved the issue, Ninth Circuit precedent that bars district courts from reviewing FCC orders controlled); Matthew N. Fulton, D.D.S., P.C. v. Enclarity, Inc., 962 F.3d 882 (6th Cir. 2020) (treating an FCC order as “persuasive,” nonbinding authority). 405 See Chevron U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 842–43 (1984); United States v. Mead Corp., 533 U.S. 218, 226–27 (2001). 406 U.S. CONST. amend. I (“Congress shall make no law . . . abridging the freedom of speech.”). Although the text of the First Amendment refers to “Congress,” it has long been understood to restrict action by the executive branch as well. See, e.g., Columbia Broad. Sys., Inc. v. Democratic Nat’l Comm., 412 U.S. 94, 160 (1973) (Douglas, J., concurring) (describing First Amendment as restricting Congress, whether “acting directly or through any of its agencies such as the FCC”); see generally, e.g., Daniel J. Hemel, Executive Action and the First Amendment’s First Word, 40 PEPP. L. REV. 601 (2013). The First Amendment applies to the states through the Fourteenth Amendment. 44 Liquormart v. Rhode Island, 517 U.S. 484, 489 n.1 (1996); U.S. CONST. amend. XIV.

Congressional Research Service 42 Section 230: An Overview

Background Principles

First Amendment Protections for Online Speech The Supreme Court has recognized that the First Amendment protects speech transmitted over the internet, saying in one case that “cyberspace,” and in particular, “social media,” is today the most important place for “the exchange of views” and other core speech activities.407 Accordingly, the Court has applied heightened scrutiny to laws that regulate online speech, particularly if they target certain types of speech based on its content.408 For example, in 1997, the Supreme Court evaluated two other provisions of the Communications Decency Act of 1996 under a strict scrutiny analysis.409 The provisions prohibited sending or displaying certain “indecent” or “patently offensive” material to minors.410 The Court said that because the law regulated “the content of speech,” the government would have to prove that the law was narrowly tailored to its goal, and ultimately ruled the provision unconstitutional.411 In the Court’s view, the broad language prohibiting “indecent” or “patently offensive” material swept in too much protected speech, encompassing “a large amount of speech that adults have a constitutional right to receive and to address to one another.”412 In addition to protecting website users when they post or read speech online, the First Amendment protects website operators when they engage in speech activities.413 Outside the context of the internet, the Supreme Court has recognized that businesses may engage in speech that receives heightened constitutional protection, for example, if they create political films414 or engage in political advocacy in the course of soliciting charitable contributions.415 Businesses will likely receive the same protections if they engage in speech online.416 The Court has also recognized that businesses engaged in speech activities generally have the right to refuse to host customers’ speech, saying that the government may violate the First Amendment if it compels “a private corporation to provide a forum for views other than its

407 Packingham v. North Carolina, 137 S. Ct. 1730, 1735 (2017); see also id. at 1737 (ruling unconstitutional a state law that prohibited convicted sex offenders from using social media, barring “access to what for many are the principal sources for knowing current events, checking ads for employment, speaking and listening in the modern public square, and otherwise exploring the vast realms of human thought and knowledge”). 408 See, e.g., Reno v. ACLU, 521 U.S. 844, 870 (1997) (“[O]ur cases provide no basis for qualifying the level of First Amendment scrutiny that should be applied to this medium.”). 409 See id. at 874. 410 Id. at 859–60. 411 Id. at 874, 879. 412 Id. at 874, 879. 413 See generally CRS Report R45650, Free Speech and the Regulation of Social Media Content, by Valerie C. Brannon. 414 Citizens United v. FEC, 558 U.S. 310, 341 (2010). 415 Riley v. Nat’l Fed’n of the Blind, 487 U.S. 781, 796 (1988). 416 See Reno, 521 U.S. at 870 (confirming that ordinary “First Amendment scrutiny” applies online).

Congressional Research Service 43 Section 230: An Overview

own.”417 This concern is heightened if the business is providing a forum for speech418 and there is a risk the user’s speech will be attributed to the business hosting it, such that the business’s decision to host the speech can be seen as an expressive choice to be associated with that speech.419 For instance, the Court has said that newspapers are engaged in constitutionally protected speech when they “exercise . . . editorial control and judgment” in decisions such as choosing what “material [will] go into a newspaper,” and has further held that the government generally may not interfere with those editorial judgments.420 A number of lower courts have extended this line of Supreme Court cases to search engines and other websites that host or present third-party content, dismissing lawsuits premised on the sites’ editorial decisions about what content to publish.421 These courts generally have not considered whether users would be likely to attribute this third-party speech to the company, but some commentators have evaluated this factor. Looking specifically at social media companies, which “are in the speech business,” one commentator has asserted that users are likely to assume that such sites choose to carry user-generated content, creating an expressive association with that speech.422 Others have questioned whether this is true, noting, for example, that social media sites have the ability to add disclaimers or otherwise disavow user speech.423 A 2016 decision by the D.C. Circuit indicates that First Amendment protection for online service providers turns on the degree of editorial judgment that those providers exercise. In U.S. Telecom

417 Pac. Gas & Elec. Co. v. Pub. Utils. Comm’n, 475 U.S. 1, 9 (1986) (plurality opinion); see also id. at 19–20 (holding that a state regulatory commission could not require a utility company to publish content in its monthly newsletter from entities who disagreed with the utility’s views); id. at 24 (Marshall, J., concurring) (“While the interference with appellant’s speech is, concededly, very slight, the State’s justification—the subsidization of another speaker chosen by the State—is insufficient to sustain even that minor burden.”). See also, e.g., Hurley v. Irish-Am. Gay, Lesbian & Bisexual Grp. of Boston, 515 U.S. 557, 576 (1995) (“[W]hen dissemination of a view contrary to one’s own is forced upon a speaker intimately connected with the communication advanced, the speaker’s right to autonomy over the message is compromised.”). 418 Manhattan Cmty. Access Corp. v. Halleck, 139 S. Ct. 1921, 1930 (2020) (“[W]hen a private entity provides a forum for speech, the private entity is not ordinarily constrained by the First Amendment because the private entity is not a state actor. The private entity may thus exercise editorial discretion over the speech and speakers in the forum.”). 419 Compare Hurley, 515 U.S. at 575 (ruling that a state could not force a parade organizer to host a specific group where the group’s “participation would likely be perceived as having resulted from the [organizer’s] . . . determination . . . that its message was worthy of presentation and quite possibly of support as well”), with Rumsfeld v. Forum for Acad. & Institutional Rights, Inc., 547 U.S. 47, 64–65 (2006) (rejecting challenge to federal funding condition requiring law schools to host military recruiters, saying the hosting decision was not “inherently expressive” and “[n]othing about recruiting suggests that law schools agree with any speech by recruiters”). 420 Miami Herald Publ’g Co. v. Tornillo, 418 U.S. 241, 258 (1974) (ruling unconstitutional a state law requiring newspapers, in certain circumstances, to publish replies to criticisms of political candidates). 421 See, e.g., La’Tiejira v. Facebook, Inc., 272 F. Supp. 3d 981, 992 (S.D. Tex. 2017); Publius v. Boyer-Vine, 237 F. Supp. 3d 997, 1008 (E.D. Cal. 2017); Zhang v. Baidu.com, Inc., 10 F. Supp. 3d 433, 443 (S.D.N.Y. 2014). 422 Berin Szóka, The First Amendment Bars Regulating Political Neutrality, Even Via Section 230, TECHDIRT (July 24, 2020), https://www.techdirt.com/articles/20200724/11372744970/first-amendment-bars-regulating-political-neutrality- even-via-section-230.shtml; accord TechFreedom, Comments on NTIA Petition for Rulemaking to Clarify Provisions of Section 230 of the Communications Act of 1934, at 43 (Sept. 2, 2020), https://techfreedom.org/wp-content/uploads/ 2020/09/NTIA-230-Petition-Comments-%E2%80%93-9.2.2020.pdf [hereinafter TechFreedom Comments on NTIA Petition]. 423 See, e.g., Daphne Keller, Who Do You Sue?, HOOVER INST., Aegis Series Paper No. 1902, at 20 (2019) (discussing whether the public does consider platforms responsible for user speech, and noting that platforms’ ability to add disclaimers or otherwise disavow user speech “arguably weaken[s] platforms’ First Amendment arguments”). Cf. Forum for Acad. & Institutional Rights, 547 U.S. at 65 (rejecting First Amendment challenge to statute requiring law schools to host military recruiters, noting “[n]othing about recruiting suggests that law schools agree with any speech by recruiters, and nothing in the [law] restricts what the law schools may say about the military’s policies”).

Congressional Research Service 44 Section 230: An Overview

Association v. FCC, the D.C. Circuit rejected a First Amendment challenge to the FCC’s 2015 net neutrality order.424 The 2015 order classified broadband internet access service providers as common carriers, subjecting them to heightened regulation, including prohibiting these providers from blocking lawful content.425 A broadband service provider argued that the rules violated its First Amendment rights by forcing providers “to transmit speech with which they might disagree.”426 The D.C. Circuit rejected this argument, concluding that there was no First Amendment issue because the FCC’s rules “affect[ed] a common carrier’s neutral transmission of others’ speech, not a carrier’s communication of its own message.”427 One critical basis for the D.C. Circuit’s conclusion was the FCC’s finding that broadband providers did not, in fact, exercise control over the content they transmitted, and instead acted “as ‘mere conduits for the messages of others, not as agents exercising editorial discretion subject to First Amendment protections.’”428 The court noted that other “entities that serve as conduits for speech produced by others” may “receive First Amendment protection,” if those entities engage in communicative activities and do not neutrally transmit “any and all users’ speech.”429 Accordingly, the court said that if a broadband provider “were to choose to exercise editorial discretion—for instance, by picking a limited set of websites to carry and offering that service as a curated internet experience—it might then qualify as a First Amendment speaker.”430 Significantly, the FCC’s 2015 order applied only to broadband providers and did not encompass, for example, so-called “edge providers” such as Google, Amazon, or Facebook, that provide content or services over the internet.431 The D.C. Circuit’s decision thus considered the First Amendment rights of only a subset of the universe of companies that could be considered providers of interactive computer services under Section 230.432 This area of the law is still developing, and as these cases demonstrate, whether any given lawsuit or regulation implicates the First Amendment by interfering with a provider’s editorial discretion will likely depend on the factual circumstances and the nuances of the lawsuit or regulation.

Section 230 Protections for Online Speech There is more precedent clarifying Section 230’s protections for moderation activities, and courts have described the law as protecting the speech of both users and providers. Section 230 arguably protects user speech by allowing providers to host user-generated content without fear of incurring liability.433 The Second Circuit said in Zeran that in enacting Section 230, Congress was, in part, responding to concerns that online providers facing potential tort liability would

424 U.S. Telecom Ass’n v. FCC, 825 F.3d 674, 740 (D.C. Cir. 2016). 425 Id. at 696. See also CRS Report R40616, The Federal Net Neutrality Debate: Access to Broadband Networks, by Patricia Moloney Figliola. 426 U.S. Telecom Ass’n, 825 F.3d at 740. 427 Id. 428 Id. at 741 (quoting In re Protecting and Promoting the Open Internet, No. 15-24, 30 FCC Rcd. 5601, 5870 (2015)) 429 Id. at 742. 430 Id. at 743. 431 See id. at 690, 695–96; see also, e.g., CRS Report R46207, Competition on the Edge of the Internet, by Clare Y. Cho (discussing edge providers). 432 See 47 U.S.C. § 230(f)(2) (defining “interactive computer service” as “any information service, system, or access software provider that provides or enables computer access by multiple users to a computer server”). 433 See, e.g., Zeran v. Am. Online, Inc., 129 F.3d 327, 330–31 (4th Cir. 1997); Ardia, supra note 72, at 386–87.

Congressional Research Service 45 Section 230: An Overview

simply prohibit or remove user content rather than litigate its legality.434 By shielding providers from that liability, Congress removed that incentive for providers to restrict user speech.435 Further, in immunizing providers’ decisions both to host and not to host user content, Section 230 can also be seen as protecting providers’ possible First Amendment rights to decide what speech to publish.436 Significantly, the way courts have interpreted Section 230(c)(1) to grant immunity for “publisher” activities seems consistent with the Supreme Court’s description of constitutionally protected “editorial” functions.437 According to the Zeran court, Congress also intended to “encourage service providers to self- regulate the dissemination of offensive material”—that is, to remove some user content.438 Granting providers immunity for their decisions to remove or restrict access to user content could operate in some tension with the goal of encouraging providers to host user speech.439 But both aspects of Section 230—providing providers with immunity for hosting user content and for restricting content—were arguably intended to ensure that the government generally would not be the entity striking the proper balance between these two goals,440 and that private parties would instead be the ones deciding whether content belonged online.441 In this sense, then, both aspects of Section 230 serve the First Amendment by shielding speech from government intervention.

434 Zeran, 129 F.3d at 330–31. 435 Id. at 331. 436 Cf., e.g., Langdon v. Google, Inc., 474 F. Supp. 2d 622, 629–31 (D. Del. 2007) (concluding plaintiff’s claims are barred by both the First Amendment and 47 U.S.C. § 230(c)(2)(A)). 437 Compare, e.g., Zeran, 129 F.3d at 330 (“[L]awsuits seeking to hold a service provider liable for its exercise of a publisher’s traditional editorial functions—such as deciding whether to publish, withdraw, postpone or alter content— are barred [by Section 230].”), with, e.g., Miami Herald Publ’g Co. v. Tornillo, 418 U.S. 241, 258 (1974) (“The choice of material to go into a newspaper, and the decisions made as to limitations on the size and content of the paper, and treatment of public issues and public officials . . . constitute the exercise of editorial control and judgment.”). 438 Zeran, 129 F.3d at 331. 439 See, e.g., Batzel v. Smith, 333 F.3d 1018, 1028 (9th Cir. 2003) (“We recognize that there is an apparent tension between Congress’s goals of promoting free speech while at the same time giving parents the tools to limit the material their children can access over the Internet. . . . [L]aws often have more than one goal in mind, and . . . it is not uncommon for these purposes to look in opposite directions. . . . Tension within statutes is often not a defect but an indication that the legislature was doing its job.”). 440 See 47 U.S.C. § 230(a)(4) (finding that the internet has “flourished, to the benefit of all Americans, with a minimum of government regulation”); id. § 230(b)(2) (stating that it is the policy of the United States “to preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation”). 441 See, e.g., 141 CONG. REC. H8470 (daily ed. Aug. 4, 1995) (statement of Rep. Christopher Cox) (“[W]e do not wish to have a Federal Computer Commission with an army of bureaucrats regulating the Internet . . . .”); id. at H8470 (statement of Rep. Joe Barton) (arguing Section 230 provides “a reasonable way to . . . help [service providers] self- regulate . . . without penalty of law”); id. at H8471 (statement of Rep. Rick White) (arguing the responsibility for “protect[ing children] from the wrong influences on the Internet” should lie with parents instead of federal government); id. at H8471 (statement of Rep. Bob Goodlatte) (“The Cox-Wyden amendment is a thoughtful approach to keep smut off the net without government censorship.”).

Congressional Research Service 46 Section 230: An Overview

Section 230 accordingly overlaps somewhat with the First Amendment. However, while Section 230 may protect some speech activities, Section 230 is not constitutionally required,442 or even coextensive with the First Amendment’s protections,443 as discussed in more detail below.444

First Amendment Issues with Reform Proposals Any legislative proposal that regulates online content may implicate the First Amendment, to the extent that it burdens protected speech activity. As currently drafted, Section 230 does not itself make any content unlawful. Instead, it governs whether interactive computer service providers and users may be subject to liability under other laws for their interactions with others’ content.445 Further, although Section 230 can be seen as speech-protective, the removal of Section 230’s statutory speech protections would not affect the scope of any constitutional speech protections. Section 230 is not constitutionally required, and Congress could repeal it without violating the First Amendment.446 Section 230 nonetheless affects constitutionally protected speech by creating government incentives for certain speech activities, and accordingly, amendments to Section 230 could implicate constitutional free speech concerns.447 However, a law is not necessarily unconstitutional merely because it affects protected speech. Courts apply a variety of different tests to determine whether government regulations implicating First Amendment interests are constitutional.448 Which analysis a court adopts depends on a variety of factors, including whether the regulation is focused primarily on speech or on conduct,449 and whether the regulation targets only certain types of speech.450 In particular, if a regulation targets speech based on its content or viewpoint, it will generally be “presumptively unconstitutional and may be justified only if the government proves that [it is] narrowly tailored to serve compelling state interests.”451

442 See, e.g., Gucci Am., Inc. v. Hall & Assocs., 135 F. Supp. 2d 409, 422 (S.D.N.Y. 2001) (“Section 230 reflects a ‘policy choice,’ not a First Amendment imperative, to immunize ISPs from defamation and other ‘tort-based lawsuits,’ driven, in part, by free speech concerns.” (quoting Zeran, 129 F.3d at 330–31)). 443 See, e.g., Eric Goldman, Why Section 230 Is Better than the First Amendment, 95 NOTRE DAME L. REV. ONLINE 33, 34 (2019) (“Section 230 provides significant and irreplaceable substantive and procedural benefits beyond the First Amendment’s free speech protections.”). 444 Infra “Comparing the Operation of First Amendment and Section 230 Protections .” 445 See 47 U.S.C. § 230. 446 See, e.g., Gucci Am., Inc., 135 F. Supp. 2d at 422. 447 See, e.g., Derek E. Bambauer, How Section 230 Reform Endangers Internet Free Speech, BROOKINGS (July 1, 2020), https://www.brookings.edu/techstream/how-section-230-reform-endangers-internet-free-speech; Keller, supra note 422, at 3; Adam Thierer & Neil Alan Chilson, FCC’s O’Rielly on First Amendment & Fairness Doctrine Dangers, FEDERALIST SOC’Y (Aug. 6, 2020), https://fedsoc.org/commentary/fedsoc-blog/fcc-s-o-rielly-on-first-amendment- fairness-doctrine-dangers. 448 See generally, e.g., CRS Report R45650, Free Speech and the Regulation of Social Media Content, by Valerie C. Brannon. 449 See, e.g., Ashcroft v. Free Speech Coal., 535 U.S, 234, 253 (2002) (“[T]he Court’s First Amendment cases draw vital distinctions between words and deeds, between ideas and conduct.”); Rumsfeld v. Forum for Acad. & Institutional Rights, Inc., 547 U.S. 47, 62 (2006) (upholding law where “the compelled speech . . . is plainly incidental to the [law’s] regulation of conduct”). 450 While certain categories of speech may be more easily regulated or even prohibited, in general, a law that targets speech because of its content will be subject to strict scrutiny, and more likely to be held unconstitutional. See generally CRS In Focus IF11072, The First Amendment: Categories of Speech, by Victoria L. Killion. 451 Reed v. Town of Gilbert, 135 S. Ct. 2218, 2226 (2015). A law will be considered content-based if it “applies to particular speech because of the topic discussed or the idea or message expressed.” Id. at 2227.

Congressional Research Service 47 Section 230: An Overview

Section 230 already contains content-based distinctions: Section 230(c)(2) extends immunity only to those providers and users restricting access to certain types of “objectionable” content,452 arguably regulating speech on the basis of its content and viewpoint.453 Any reform proposals that add to the current list of types of content in Section 230(c)(2) could create additional content- or viewpoint-based distinctions,454 potentially triggering heightened scrutiny under prevailing Supreme Court precedent.455 Other bills from the 116th Congress would have limited providers’ editorial discretion by extending immunity only to providers that moderate content in specific types of ways.456 For example, some proposals would have required providers to moderate in a viewpoint-neutral manner to qualify for Section 230 protections457—and at least one bill would have extended the same requirement to users.458 To the extent that these proposals “draw[] distinctions based on the message a speaker conveys,” or based on the “communicative content” of the speech, they might also be subject to heightened scrutiny if enacted and challenged in court.459 No court has given significant consideration to the constitutionality of Section 230 in its current form, making it difficult to say definitively how a court would view reform proposals that build on the law’s current structure.460 Some have argued that because Congress was not required to

452 47 U.S.C. § 230(c)(2) (providing immunity to providers and users for certain decisions to restrict access to “material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally protected”). 453 E.g., Eric Goldman, Sen. Graham Cares More About Trolls than Section 230 (Comments on Online Content Policy Modernization Act), TECH. & MKTG. L. BLOG (Sept. 30, 2020), https://blog.ericgoldman.org/archives/2020/09/sen- graham-cares-more-about-trolls-than-section-230-comments-on-online-content-policy-modernization-act.htm. See also Matal v. Tam, 137 S. Ct. 1744, 1763 (2020) (plurality opinion) (“Giving offense is a viewpoint.”). 454 See, e.g., Online Freedom and Viewpoint Diversity Act, S. 4534, 116th Cong. § 2 (2020) (replacing “otherwise objectionable” in Section 230(c)(2) with “promoting self-harm, promoting terrorism, or unlawful”); Stop the Censorship Act of 2020, H.R. 7808, 116th Cong. § 2 (2020) (replacing “otherwise objectionable” in Section 230(c)(2) with “unlawful, or that promotes violence or terrorism”). But cf. Stop the Censorship Act, H.R. 4027, 116th Cong. § 2 (2019) (replacing entire list of adjectives in Section 230(c)(2) with “unlawful”). 455 See Reed, 135 S. Ct. at 2226. 456 See, e.g., BAD ADS Act, S. 4337, 116th Cong. § 2 (2020) (limiting application of Section 230(c) if a covered provider displays “behavioral advertising”); Biased Algorithm Deterrence Act of 2019, H.R. 492, 116th Cong. § 2 (2019) (providing that “a social media service that displays user-generated content in an order other than chronological order . . . , if for a reason other than to restrict access to . . . material described in paragraph (2)(A) or to carry out the direction of the user that generated such content, shall be treated as a publisher . . . of such content”). 457 See, e.g., CASE-IT Act, H.R. 8719, 116th Cong. § 2 (2020) (providing that Sections 230(c)(1) and (c)(2)(A) will not apply to certain providers if the provider “makes content moderation decisions pursuant to policies or practices that are not reasonably consistent with the First Amendment to the Constitution,” treating these providers as equivalent to state actors); Ending Support for Internet Censorship Act, S. 1914, 116th Cong. § 2 (2019) (providing that Section 230(c) will not apply to larger providers unless the FTC has certified that “the company does not moderate information . . . in a manner that is biased against a political party, political candidate, or political viewpoint”). 458 Stopping Big Tech’s Censorship Act, S. 4062, 116th Cong. § 2 (2020) (providing that Section 230(c)(2)(A) will apply only if a provider or user, among other requirements, acts “in a viewpoint-neutral manner”). 459 Reed, 135 S. Ct. at 2226. 460 Cf. Green v. Am. Online (AOL), 318 F.3d 465, 470 (3rd Cir. 2003) (rejecting plaintiff’s claim that Section 230(c)(2) violates the First Amendment because it allows providers to restrict constitutionally protected material by noting that the provision did “not require” the provider to “restrict speech”). In a document defending its petition for rulemaking, NTIA argued that the Supreme Court has previously “upheld federal compelled speech in exchange for liability protections,” citing Farmers Educational & Cooperative Union of America v. WDAY, Inc., 360 U.S. 525 (1959). NTIA, Reply Comments on NTIA Petition for Rulemaking to Clarify Provisions of Section 230 of the Communications Act of 1934, at 37 (Sept. 17, 2020), https://ecfsapi.fcc.gov/file/1091762203541/NTIA%20Reply%20Comments%20in% 20RM%20No.%2011862.pdf. In that case, the Court held that a federal provision prohibiting licensed broadcasters from “censor[ing]” certain statements of political candidates should be read to implicitly grant stations immunity from

Congressional Research Service 48 Section 230: An Overview

grant this immunity, it can restrict or condition Section 230 immunity without raising any constitutional concerns.461 Even if the government places conditions on Section 230, they argue, the government has not compelled providers or users to moderate speech in any particular way, but has merely incentivized it to do so.462 Other commentators have argued such speech-based limits on Section 230 immunity would run afoul of Supreme Court precedent prohibiting unconstitutional conditions on benefits.463 In other contexts, the Supreme Court has recognized that denying a benefit “to claimants who engage in certain forms of speech is in effect to penalize them for such speech,” and can have the same “deterrent effect” as a more direct regulation.464 Under the unconstitutional conditions doctrine, which has largely—but not solely465—been developed in the context of grant programs,466 the government “may not deny a benefit to a person on a basis that infringes his constitutionally protected interests.”467 While the government may impose conditions that ensure funds “will be used only to further the purposes of a grant,”468 it may violate the First Amendment if it uses a grant program to impose restrictions on private speech.469 By contrast, the Supreme Court has held that when the government is using a grant program to recruit “private entities to convey a governmental message,” it may impose content- and viewpoint-based restrictions on funded speech.470 liability for defamatory statements made in those broadcasts. Farmers Educ. & Coop. Union of Am., 360 U.S. at 526, 535 (discussing 47 U.S.C. § 315). The opinion did not involve a First Amendment analysis. 461 See, e.g., Craig Parshall, Big Tech and the Whole First Amendment, FEDERALIST SOC’Y (Aug. 14, 2020), https://fedsoc.org/commentary/fedsoc-blog/big-tech-and-the-whole-first-amendment. 462 See, e.g., Free State Foundation, Comments on NTIA Petition for Rulemaking to Clarify Provisions of Section 230 of the Communications Act of 1934, at 7 (Sept. 17, 2020), https://ecfsapi.fcc.gov/file/10917275329599/Section% 20230%20Petition%20FSF%20Reply%20Comments%20-%20Final.pdf [hereinafter Free State Foundation Comments on NTIA Petition]. 463 See, e.g., Elliot Harmon, The Online Content Policy Modernization Act Is an Unconstitutional Mess, ELECTRONIC FRONTIER FOUNDATION (Oct. 1, 2020), https://www.eff.org/deeplinks/2020/10/online-content-policy-modernization- act-unconstitutional-mess; Mark Joseph Stern, Josh Hawley Wants to Stop Internet Censorship by Censoring the Internet, SLATE (June 19, 2019), https://slate.com/technology/2019/06/josh-hawley-section-230-cda-internet-speech- conservatives.html. 464 Speiser v. Randall, 357 U.S. 513, 518 (1958); see also id. at 529 (concluding that a California provision requiring veterans seeking a property tax exemption to swear a loyalty oath was unconstitutional because it placed the burden of proof on the claimants). Cf., e.g., Free State Foundation Comments on NTIA Petition, supra note 462, at 8 (acknowledging that conditions on Section 230 immunity burdening a platform’s ability to host or moderate could be relevant factors in a First Amendment challenge). 465 See, e.g., Frost & Frost Trucking Co. v. R.R. Comm’n, 271 U.S. 583, 593–94 (1926) (holding that a state could not place conditions on permits that would “require the relinquishment of constitutional rights”). 466 Cf. Matal v. Tam, 137 S. Ct. 1744, 1760–61 (2020) (plurality opinion) (describing unconstitutional condition cases as involving “cash subsidies or their equivalent”). 467 Perry v. Sindermann, 408 U.S. 593, 597 (1972) (holding that a public university could not place a condition on employment that violated a person’s free speech rights). 468 Rust v. Sullivan, 500 U.S. 173, 198 (1991) (upholding a federal grant condition prohibiting health programs receiving federal funding from encouraging the use of abortion). 469 See Agency for Int’l Dev. v. All. for Open Soc’y Int’l, Inc., 570 U.S. 205, 208, 218 (2013) (holding that a federal condition requiring funding recipients to have “a policy explicitly opposing prostitution and sex trafficking” was unconstitutional because it limited the recipients’ speech outside the bounds of the federal program); Legal Servs. Corp. v. Velazquez, 531 U.S. 533, 542, 548–49 (2001) (holding that a federal condition prohibiting funds from being used for legal representation involving an effort to amend welfare law was unconstitutional, where the program “was designed to facilitate private speech, not to promote a governmental message”). 470 Rosenberger v. Rector & Visitors of the Univ. of Va., 515 U.S. 819, 833–34 (1995); see also Matal v. Tam, 137 S. Ct. 1744, 1768 (2020) (Kennedy, J., concurring) (“[T]he Court’s precedents have recognized just one narrow situation

Congressional Research Service 49 Section 230: An Overview

Section 230 grants a legal benefit, in the form of immunity. As discussed above, when providers and users choose to distribute or restrict others’ content, they can be seen as engaged in constitutionally protected speech.471 Accordingly, because Section 230 immunity contains certain conditions related to these speech decisions, the unconstitutional conditions doctrine may be seen as an appropriate framework to analyze the law’s constitutionality, or the constitutionality of any reform proposals that would similarly condition immunity.472 Accordingly, some commentators have suggested that proposals that would “forc[e]” private actors to change their decisions about what speech to host or distribute would create an unconstitutional condition by causing them to “surrender [their] First Amendment rights” to qualify for the benefit of legal immunity.473 As four members of the Supreme Court recognized in 2017’s Matal v. Tam, though, the Court’s precedent applying the unconstitutional conditions doctrine is not easily extended outside the context of programs that provide “cash subsidies or their equivalent.”474 For example, it is difficult to apply cases asking whether a speech restriction serves “the purposes of a grant”475 to review conditions on Section 230 immunity. However, the Court has recognized constitutional limitations on the government’s ability to condition legal privileges outside the context of grant programs—for example, in the context of permits.476 Accordingly, even in the context of broadcast media, although the Supreme Court has recognized that the government has greater leeway to regulate this particular medium by placing conditions on broadcast licenses,477 the Court has still struck down regulations that unduly interfere with licensees’ ability to express their own “editorial opinion.”478

in which viewpoint discrimination is permissible: where the government itself is speaking or recruiting others to communicate a message on its behalf.”). But cf. Nat’l Endowment for the Arts v. Finley, 524 U.S. 569, 580–87 (1998) (rejecting a variety of arguments claiming that a program requiring the federal grantor to take into account “general standards of decency” discriminated on the basis of viewpoint). 471 See, e.g., Publius v. Boyer-Vine, 237 F. Supp. 3d 997, 1008 (E.D. Cal. 2017) (concluding that a website owner had “a First Amendment right to distribute and facilitate protected speech on the site”). 472 See, e.g., Edwin Lee, Conditioning Section 230 Immunity on Unbiased Content Moderation Practices as an Unconstitutional Condition, 2020 U. ILL. J.L. TECH. & POL’Y 457, 466 (2020); Mark MacCarthy, Some Reservations About a Consistency Requirement for Social Media Content Moderation Decisions, FORBES (July 29, 2020) https://www.forbes.com/sites/washingtonbytes/2020/07/29/some-reservations-about-a-consistency-requirement-for- social-media-content-moderation-decisions/?sh=524e846276d7. Cf., e.g., Cathy Gellis, Section 230 Isn’t a Subsidy; It’s a Rule of Civil Procedure, TECHDIRT (Dec. 29, 2020), https://www.techdirt.com/articles/20201229/12003745970/ section-230-isnt-subsidy-rule-civil-procedure.shtml (arguing that Section 230 is more similar to a rule of civil procedure than “some sort of tangible prize the government hands out selectively”). 473 TechFreedom Comments on NTIA Petition, supra note 422, at 37; accord Internet Association, Comments Opposing the NTIA’s Petition for Rulemaking, at 50–51 (Sept. 2, 2020), https://ecfsapi.fcc.gov/file/10902184309650/ IA%20Comments%20to%20FCC%20on%20NTIA%20Petition%20re%20Section%20230_v2.pdf [hereinafter Internet Association Comments on NTIA Petition]. 474 Matal v. Tam, 137 S. Ct. 1744, 1761 (2020) (plurality opinion). 475 See Rust v. Sullivan, 500 U.S. 173, 198 (1991). 476 See, e.g., Frost & Frost Trucking Co. v. R.R. Comm’n, 271 U.S. 583, 593–94 (1926). 477 Red Lion Broad. Co. v. FCC, 395 U.S. 367, 400–01 (1969) (rejecting a constitutional challenge to a regulation requiring broadcasters to carry certain content “[i]n view of the scarcity of broadcast frequencies, the Government’s role in allocating those frequencies, and the legitimate claims of those unable without governmental assistance to gain access to those frequencies for expression of their views”). See also Reno v. ACLU, 521 U.S. 844, 868 (1997) (noting that these “special justifications for regulation of the broadcast media . . . are not applicable to other speakers” and specifically rejecting the idea that the internet should receive similar special First Amendment treatment). 478 FCC v. League of Women Voters, 468 U.S. 364, 381 (1984). This particular case involved a condition on a grant program administered by the Corporation for Public Broadcasting, but the condition was analyzed under the constitutional rubric that applies to broadcast licensees. See id. at 377–78.

Congressional Research Service 50 Section 230: An Overview

Further, although the plurality opinion in Tam declined to apply the unconstitutional conditions doctrine outside the context of a cash subsidy,479 the decision nonetheless suggests that viewpoint- based conditions to Section 230 may pose constitutional problems.480 In Tam, the Court held that a federal law prohibiting the registration of disparaging trademarks was unconstitutional under the First Amendment, saying that “[s]peech may not be banned on the ground that it expresses ideas that offend.”481 Like Section 230, the federal trademark law did not directly prohibit disparaging speech, but merely limited the benefits of federal trademark registration.482 Ultimately, the Court ruled the law unconstitutional, saying that determining whether a mark was so “offensive” that it could not be registered entailed impermissible viewpoint discrimination.483 The Court rejected arguments claiming that the viewpoint discrimination was acceptable because trademarks can be seen as government speech, saying instead that trademarks are private speech.484 Because Section 230 provides immunity for private speech activities and similarly cannot be framed as advancing a government message, Tam could suggest that viewpoint-based conditions on Section 230 immunity are unconstitutional.485 Ultimately, it is difficult to say definitively how a court would analyze a First Amendment challenge to a limit or condition on Section 230 immunity, although Supreme Court precedent suggests that laws that draw distinctions based on the content or viewpoint of speech may be subject to heightened scrutiny, even in the context of a law that merely disfavors, rather than prohibits, certain speech.486 However, the fact that any given Section 230 reform proposal does not directly prohibit or compel speech would certainly be a relevant factor in the First Amendment analysis.487 On the other hand, if a Section 230 reform proposal more directly requires providers or users to distribute or restrict content,488 it may raise heightened First Amendment concerns. Content-neutral proposals would likely be evaluated under a more lenient standard and would be more likely to be upheld against a First Amendment challenge.489 Specifically, content-neutral

479 Matal v. Tam, 137 S. Ct. 1744, 1761 (2020) (plurality opinion). 480 See id. at 1763; id. at 1765 (Kennedy, J., concurring). 481 Id. at 1751 (majority opinion). 482 See id. at 1753 (discussing the legal rights and benefits conferred by registration). 483 Id. at 1763 (plurality opinion); see also id. at 1765 (Kennedy, J., concurring) (“The law . . . reflects the Government’s disapproval of a subset of messages it finds offensive. This is the essence of viewpoint discrimination.”). 484 Id. at 1760 (majority opinion). 485 See id.; see also Legal Servs. Corp. v. Velazquez, 531 U.S. 533, 548–49 (2001) (“Where private speech is involved, even Congress’ antecedent funding decision cannot be aimed at the suppression of ideas thought inimical to the Government’s own interest.”). 486 See, e.g., Matal, 137 S. Ct. at 1751 (majority opinion); United States v. Playboy Entm’t Group, 529 U.S. 803, 809, 827 (2000) (holding that federal statute restricting the availability of “sexually explicit [cable] channel[s]” discriminated on the basis of content and was unconstitutional under a strict scrutiny analysis). Cf., e.g., Backpage.com, LLC v. Dart, 807 F.3d 229, 230 (7th Cir. 2015) (granting an injunction against a sheriff who “embarked on a campaign intended to crush Backpage’s adult section” by sending letters to credit card companies “demanding” that they “prohibit the use of their credit cards to purchase any ads on Backpage”). 487 See, e.g., Free State Foundation Comments on NTIA Petition, supra note 462, at 7–8. 488 See, e.g., See Something, Say Something Online Act of 2020, S. 4758, 116th Cong. § 5 (2020) (amending Section 230 to include an affirmative requirement for providers to “take reasonable steps to prevent or address unlawful users [sic] of the service through the reporting of suspicious transmissions”); CASE-IT Act, H.R. 8719, 116th Cong. § 2 (2020) (creating a new private right of action allowing content providers to sue service providers that fail “to make content moderation decisions pursuant to policies or practices that are reasonably consistent with the First Amendment to the Constitution”). 489 See, e.g., Univ. City Studios, Inc. v. Corley, 273 F.3d 429, 451, 454 (2d Cir. 2001) (rejecting a First Amendment

Congressional Research Service 51 Section 230: An Overview

laws that regulate speech are subject to intermediate scrutiny, which asks whether the restriction is “narrowly tailored to serve a significant governmental interest” and “leave[s] open ample alternative channels for communication of the information.”490 Further, Congress may be able to target certain limited categories of speech that the Supreme Court has historically recognized can be regulated more freely, such as obscenity or fraud, without triggering heightened scrutiny.491

Comparing the Operation of First Amendment and Section 230 Protections Besides the constitutionality of Section 230’s immunity provisions and proposed reforms, another relevant issue is the extent to which the First Amendment might prevent liability for hosting content. The scope of First Amendment protections is important to understand the potential consequences of Section 230 reforms. For example, FOSTA both created a new federal criminal offense and created new exceptions to Section 230 immunity.492 The new criminal offense, which prohibits operating an interactive computer service “with the intent to promote or facilitate the prostitution of another person,”493 has been challenged on constitutional grounds.494 The D.C. Circuit ruled in January 2020 that one such challenge should be allowed to proceed, concluding that the statute could apply to plaintiffs engaging in advocacy or educational activities that might be protected by the First Amendment.495 By contrast, in January 2021, a Texas federal district court rejected a criminal defendant’s First Amendment challenge to the provision.496 These cases will likely affect not only the government’s ability to enforce this federal criminal law, but will also be relevant for courts determining whether providers and users can be held liable under the FOSTA exceptions to Section 230 immunity. Namely, even though Section 230 no longer bars state criminal prosecutions that track this new criminal offense,497 courts might nonetheless conclude that the First Amendment prevents prosecution.498 In a variety of legal contexts, courts have suggested that the First Amendment imposes a heightened standard of liability, such as requiring proof of a higher level of intent, before speech “distributors” such as bookstores and newsstands can be punished for circulating unlawful content.499 And even in the context of lawsuits against publishers such as newspapers or challenge to a court order enforcing the Digital Millennium Copyright Act because the restriction targeted the “functional,” “nonspeech” aspects of computer code, and was accordingly content neutral). 490 Ward v. Rock Against Racism, 491 U.S. 781, 791 (1989) (quoting Clark v. Cmty. for Creative Non-Violence, 468 U.S. 288, 293 (1984)). 491 See United States v. Stevens, 559 U.S. 460, 468–69 (2010) (listing these categories as including obscenity, defamation, fraud, incitement, and speech integral to criminal conduct); R.A.V. v. City of St. Paul, 505 U.S. 377, 383– 84 (1992) (discussing the government’s ability to regulate these categories). 492 Allow States and Victims to Fight Online Sex Trafficking Act of 2017, Pub. L. No. 115-164, §§ 3–4, 132 Stat. 1253, 1253–54 (2018). 493 18 U.S.C. § 2421A. 494 See Woodhull Freedom Found. v. United States, 948 F.3d 363, 367 (D.C. Cir. 2020); United States v. Martono, No. 3:20-CR-00274-N-1, 2021 WL 39584, at *1 (N.D. Tex. Jan. 5, 2021). 495 Woodhull Freedom Found., 948 F.3d at 372–73. 496 Martono, 2021 WL 39584, at *1 (concluding the law was not unconstitutionally overbroad or vague). 497 47 U.S.C. § 230(e)(5)(C) (providing that Section 230 will not “impair or limit . . . any charge in a criminal prosecution brought under State law if the conduct underlying the charge would constitute a violation of [18 U.S.C. § 2421A] and promotion or facilitation of prostitution is illegal in the jurisdiction where the defendant’s promotion or facilitation of prostitution was targeted”). 498 Cf., e.g., State v. Melchert-Dinkel, 844 N.W.2d 13, 23–24 (Minn. 2014) (concluding that a state criminal law prohibiting advising or encouraging another to commit suicide violates the First Amendment). 499 See, e.g., Smith v. California, 361 U.S. 147, 155 (1959) (holding that a law imposing criminal penalties on

Congressional Research Service 52 Section 230: An Overview

magazines, courts have sometimes imposed heightened standards where the liability is premised on speech.500 Consequently, some commentators have argued that even if Section 230 were repealed, the First Amendment would nonetheless continue to prevent liability premised on hosting or distributing speech.501 Although the Constitution likely would preclude civil or criminal liability in some circumstances, the protections of the First Amendment are likely not coextensive with Section 230 immunity.502 First, while Section 230 provides a complete bar to liability for covered activities, the First Amendment may merely impose a heightened standard of liability if a lawsuit implicates protected speech.503 One illustration comes from the New York rulings described above that considered whether early online platforms hosting message boards could be held liable for defamatory statements posted by users.504 In Cubby, the federal trial court concluded that CompuServe should be treated as a distributor for purposes of analyzing the defamation claim.505 Accordingly, the court ruled that the plaintiff had to meet a heightened standard and prove that CompuServe “knew or had reason to know of the allegedly defamatory . . . statements.”506 While the trial court ultimately concluded that the plaintiff had not met this standard and dismissed the defamation claim,507 it was theoretically possible for the plaintiff to prove the claim by submitting sufficient evidence of CompuServe’s knowledge. By contrast, courts have ruled that Section 230 will bar a claim against a provider that merely publishes a defamatory statement regardless of whether the provider actually knew about the statement.508 Accordingly, while the types of heightened standards required by the First Amendment likely would lead courts to dismiss some lawsuits premised on speech, plaintiffs with sufficient proof may be able to overcome those standards in circumstances where Section 230 would have barred the suit. However, a few trial bookstores that possess obscene material was unconstitutional under the First Amendment because it did not include any element of scienter, or knowledge); Cubby, Inc. v. CompuServe, Inc., 776 F. Supp. 135, 139 (S.D.N.Y. 1991) (requiring proof of knowledge before a distributor may be held liable for defamation). See also, e.g., Bantam Books, Inc. v. Sullivan, 372 U.S. 58, 70 (1963) (holding that a state commission violated the First Amendment by sending book publishers notices threatening punishment under state obscenity laws, characterizing the scheme as a system of prior administrative restraints that was impermissible because it lacked sufficient procedural safeguards). 500 See, e.g., N.Y. Times Co. v. Sullivan, 376 U.S. 254, 279–80 (1964) (requiring a showing of “actual malice” before a “public official” may recover damages from a newspaper for a defamatory statement “relating to his official conduct”); Braun v. Soldier of Fortune Magazine, Inc., 968 F.2d 1110, 1114 (11th Cir. 1992) (ruling that a magazine could be held liable for negligently publishing an advertisement “only if the advertisement on its face would have alerted a reasonably prudent publisher to the clearly identifiable unreasonable risk of harm”). 501 See, e.g., Note, Section 230 as First Amendment Rule, 131 HARV. L. REV. 2027, 2028 (2018); cf. Brent Skorup & Jennifer Huddleston, The Erosion of Publisher Liability in American Law, Section 230, and the Future of Online Curation, 72 OKLA. L. REV. 635, 637 (2020) (arguing that in the area of defamation law, “First Amendment considerations would likely lead courts to a § 230-like liability protection,” but noting differences in the two regimes). 502 See generally, e.g., Goldman, Why Section 230 Is Better than the First Amendment, supra note 443 (discussing ways Section 230 offers more protection, both substantive and procedural, than the First Amendment). 503 See, e.g., id. at 38–39 (noting that “sufficient scienter can override” First Amendment protections in defamation cases, but Section 230 “moot[s] inquiries into defendants’ scienter”). 504 Compare Stratton Oakmont, Inc. v. Prodigy Servs. Co., No. 31063/94, 1995 WL 323710, at *4 (N.Y. Sup. Ct. May 24, 1995) (concluding Prodigy should be treated as a publisher for purposes of defamation liability), with Cubby, Inc. v. CompuServe, Inc., 776 F. Supp. 135, 140–41 (S.D.N.Y. 1991) (concluding CompuServe should be treated as a distributor for purposes of defamation liability). 505 Cubby, Inc., 776 F. Supp. at 140. 506 Id. at 140–41. 507 Id. at 141. 508 See, e.g., Zeran v. Am. Online, Inc., 129 F.3d 327, 334 (4th Cir. 1997) (concluding Section 230 barred claim that provider could be held liable for defamation as a distributor with knowledge of the statement).

Congressional Research Service 53 Section 230: An Overview

courts have concluded that the First Amendment completely immunizes websites from certain civil claims without suggesting that some heightened standard applies—similar to the current regime under Section 230.509 Section 230 also provides complete immunity for “publisher” activities absent an inquiry into whether the underlying content is constitutionally protected, meaning that Section 230 likely protects hosting at least some speech that the First Amendment does not protect.510 As discussed above, the inquiry into whether a service provider or user has engaged in “publisher” activities may overlap with constitutional protections for “editorial” activity,511 but Section 230 nonetheless does not require a court to investigate whether First Amendment activity has occurred. Accordingly, Section 230 provides greater certainty for service providers and users that distributing or restricting others’ speech will be protected from liability, without having to consider whether a court would conclude the speech is constitutionally protected.512 In at least some cases, courts may dismiss a lawsuit against a provider on Section 230 grounds at an early stage in the litigation based on the allegations alone.513 Whether early dismissal is warranted, however, will depend on the elements of the claim, the factual circumstances, and the particulars of any Section 230 or First Amendment defense. For example, as discussed above, allegations that a provider acted in bad faith have prevented providers from obtaining early dismissal under Section 230(c)(2)(A).514 Accordingly, while the First Amendment might prevent some claims premised on decisions to host or restrict others’ speech, its protections are likely less extensive than the current scope of Section 230 immunity.

Author Information

Valerie C. Brannon Eric N. Holmes Legislative Attorney Legislative Attorney

509 E.g., Zhang v. Baidu.com, Inc., 10 F. Supp. 3d 433, 443 (S.D.N.Y. 2014); Langdon v. Google, Inc., 474 F. Supp. 2d 622, 629–30 (D. Del. 2007); Search King, Inc. v. Google Tech., Inc., No. CIV-02-1457-M, 2003 WL 21464568, at *3– 4 (W.D. Okla. May 27, 2003). 510 See, e.g., 47 U.S.C. § 230(c)(1) (providing that a service provider or user may not be treated as a “publisher” of another’s content); id. § 230(c)(2) (extending immunity for decisions to restrict certain material, “whether or not such material is constitutionally protected”). 511 See supra note 437. 512 See, e.g., Goldman, Why Section 230 Is Better than the First Amendment, supra note 443, at 42–43. 513 See id. at 39–40; accord Gellis, supra note 472. 514 See supra note 198.

Congressional Research Service 54 Section 230: An Overview

Disclaimer This document was prepared by the Congressional Research Service (CRS). CRS serves as nonpartisan shared staff to congressional committees and Members of Congress. It operates solely at the behest of and under the direction of Congress. Information in a CRS Report should not be relied upon for purposes other than public understanding of information that has been provided by CRS to Members of Congress in connection with CRS’s institutional role. CRS Reports, as a work of the United States Government, are not subject to copyright protection in the United States. Any CRS Report may be reproduced and distributed in its entirety without permission from CRS. However, as a CRS Report may include copyrighted images or material from a third party, you may need to obtain the permission of the copyright holder if you wish to copy or otherwise use copyrighted material.

Congressional Research Service R46751 · VERSION 1 · NEW 55